This consolidated appeal is from the judgments of the United States District Court for the Southern District of New York (Pauley, /.) dismissing two complaints for lack of subject matter jurisdiction under the Alien Tort Statute, 28 U.S.C. § 1350 (“ATS”), and in the alternative, on the ground of forum non conveniens. Plaintiffs-Appellants Rabi Abdullah! and other Nigerian children and their guardians sued Defendant-Appellee Pfizer, Inc. under the ATS (“the Abdullahi action”). They alleged that Pfizer violated a customary international law norm prohibiting involuntary medical experimentation on humans when it tested an experimental antibiotic on children in Nigeria, including themselves, without their consent or knowledge. Plaintiffs-Appellants Ajudu Ismaila Adamu and others, also children and their guardians who were part of Pfizer’s Nigerian drug experiment, brought a similar action against Pfizer, alleging violations of the ATS, the Connecticut Unfair Trade Practices Act (“CUTPA”), and the Connecticut Products Liability Act (“CPLA”) (“the Adamu action”). Pfizer moved to dismiss both actions for lack of subject matter jurisdiction *169and on the basis of forum non coveniens. The district court granted the motions and both sets of plaintiffs have appealed.

As explained below, we conclude: (1) that the district court incorrectly determined that the prohibition in customary international law against nonconsensual human medical experimentation cannot be enforced through the ATS; (2) that changed circumstances in Nigeria since the filing of this appeal require re-examination of the appropriate forum, albeit on the basis of a legal analysis different from that employed by the district court; and (3) that the district court incorrectly applied Connecticut’s choice of law rules in the Adamu action. Consequently, we reverse and remand the cases to the district court for further proceedings.

BACKGROUND

A. Pfizer’s Trovan Test in Nigeria

On review of a district court’s grant of a motion to dismiss, we assume as true the facts alleged in the complaints, construing them in the light most favorable to the appellants. See Vietnam Ass’n for Victims of Agent Orange v. Dow Chem. Co., 517 F.3d 104, 115 (2d Cir. 2008). The central events at issue in these cases took place in 1996, during an epidemic of bacterial meningitis in northern Nigeria.1 The appellants allege that at that time, Pfizer, the world’s largest pharmaceutical corporation, sought to gain the approval of the U.S. Food and Drug Administration (“FDA”) for the use on children of its new antibiotic, Trovafloxacin Mesylate, marketed as “Trovan.” They contend that in April 1996, Pfizer, dispatched three of its American physicians to work with four Nigerian doctors to experiment with Trovan on children who were patients in Nigeria’s Infectious Disease Hospital (“IDH”) in Kano, Nigeria. Working in concert with Nigerian government officials, the team allegedly recruited two hundred sick children who sought treatment at the IDH and gave half of the children Trovan and the other half Ceftriaxone, an FDA-approved antibiotic the safety and efficacy of which was well-established. Appellants contend that Pfizer knew that Trovan had never previously been tested on children in the form being used and that animal tests showed that Trovan had life-threatening side effects, including joint disease, abnormal cartilage growth, liver damage, and a degenerative bone condition. Pfizer purportedly gave the children who were in the Ceftriaxone control group a deliberately low dose in order to misrepresent the effectiveness of Trovan in relation to Ceftriaxone. After approximately two weeks, Pfizer allegedly concluded the experiment and left without administering follow-up care. According to the appellants, the tests caused the deaths of eleven children, five of whom had taken Trovan and six of whom had taken the lowered dose of Ceftriaxone, and left many others blind, deaf, paralyzed, or brain-damaged.

Appellants claim that Pfizer, working in partnership with the Nigerian government, failed to secure the informed consent of either the children or their guardians and specifically failed to disclose or explain the experimental nature of the study or the serious risks involved. Although the treatment protocol required the researchers to offer or read the subjects documents requesting and facilitating their informed consent, this was allegedly not done in *170either English or the subjects’ native language of Hausa. The appellants also contend that Pfizer deviated from its treatment protocol by not alerting the children or their guardians to the side effects of Trovan or other risks of the experiment, not providing them with the option of choosing alternative treatment, and not informing them that the non-governmental organization Médecins Sans Frontiéres (Doctors Without Borders) was providing a conventional and effective treatment for bacterial meningitis, free of charge, at the same site.2

The appellants allege that, in an effort to rapidly secure FDA approval, Pfizer hastily assembled its test protocol at its research headquarters in Groton, Connecticut, and requested and received permission to proceed from the Nigerian government in March 1996. At the time, Pfizer also claimed to have secured approval from an IDH ethics committee. Appellants allege, however, that the March 1996 approval letter was backdated by Nigerian officials working at the government hospital well after the experiments had taken place and that at the time the letter was purportedly written, the IDH had no ethics committee.3 Appellants also contend that the experiments were condemned by doctors, including one on Pfizer’s staff at the time of the Kano trial.

In 1998, the FDA approved Trovan for use on adult patients only. After reports of liver failure in patients who took Trovan, its use in America was eventually restricted to adult emergency care. In 1999, the European Union banned its use.

B. The Proceedings Below

In August 2001, the Abdullahi plaintiffs sued Pfizer under the ATS, alleging that the experiments violated international law. In September 2002, the district court granted Pfizer’s motion to dismiss the Abdullahi claims on the ground oí forum non conveniens, conditioned on Pfizer’s consent to litigation in Nigeria. Abdullahi v. Pfizer, Inc., No. 01 Civ. 8118(WHP), 2002 WL 31082956, at *12 (S.D.N.Y. Sept. 17, 2002) (“Abdullahi I ”). It found that Nigeria was an adequate alternative forum despite plaintiffs’ contentions about corruption in the Nigerian court system. Id. at *8-10. The district court denied Pfizer’s motion to dismiss under Rule 12(b)(6), Fed.R.Civ.P., concluding that the plaintiffs adequately alleged that Pfizer’s collusion with the Nigerian government made it a state actor. Id. at *5-6.

Meanwhile, another group of children and guardians involved in the Trovan experiment sued in the Federal High Court in Kano, alleging claims under Nigerian law. That case, Zango v. Pfizer International, Inc., [2001] Suit No. FHC/ *171K/CS/204/2001 (Nigeria), was dismissed in 2003 after plaintiffs voluntarily discontinued the suit following the removal from the bench of the first judge assigned to the action and the second judge’s decision to decline jurisdiction for personal reasons. Abdullahi v. Pfizer, Inc., No. 01 Civ. 8118(WHP), 2005 WL 1870811, at *5 (S.D.N.Y. Aug. 9, 2005) (“Abdullahi III”). On appeal to this Court from the district court’s dismissal in Abdullahi I, the Abdullahi appellants argued that the dismissal of the Zango litigation was a result of rampant corruption, which indicated that the Nigerian judicial system could not provide an adequate alternative forum for their action. Given an inconclusive record regarding the events leading to the dismissal of the Zango lawsuit, we vacated the judgment and remanded for further fact-finding on forum,' non conveniens. See Abdullahi v. Pfizer, Inc., 77 Fed.Appx. 48, 53 (2d Cir. 2003) (summary order) (“Abdullahi II ”).

In November 2002, following the dismissal of the Zango lawsuit, a number of the Zango plaintiffs filed the Adamu action. They alleged that in planning the Trovan experiment in Connecticut and in conducting the tests in Nigeria without informed,consent, Pfizer violated the CUT-PA, the CPLA, and the ATS. Eventually, the Adamu action was transferred to the Southern District of New York and consolidated with the Abdullahi action. Pfizer then moved to dismiss both cases for failure to state a claim under the ATS and on the basis of forum non conveniens. It also moved to dismiss in Adamu on the ground that Connecticut choice of law principles require the application of Nigerian law, which bars suit under CUTPA and the CPLA.

The district court granted the motions. See Abdullahi III, 2005 WL 1870811; Adamu v. Pfizer, Inc., 399 F.Supp.2d 495 (S.D.N.Y. 2005). In Abdullahi III, Judge Pauley held that while “[pjlaintiffs correctly state that non-consensual medical experimentation violates the law of nations and, therefore, the laws of the United States,” they failed to identify a source of international law that “provide[s] a proper predicate for jurisdiction under the ATS.” 2005 WL 1870811, at *9, 14. Noting that “a decision to create a private right of action is one better left to legislative judgment in the great majority of cases,” he concluded that “[a] cause of action for Pfizer’s failure to get any consent, informed or otherwise, before performing medical experiments on the subject children would expand customary international law far beyond that contemplated by the ATS.” Id. at *13-14 (internal quotation marks omitted).

With regard to the forum non conveniens analysis, the district court declined to accept plaintiffs’ submissions concerning Pfizer’s alleged bribery of Nigerian officials on the ground that they were not based on personal knowledge. Id. at *16-17. Finding that the plaintiffs had failed to submit specific evidence that the Nigerian judiciary would be biased against its own citizens in an action against Pfizer, the district court alternatively held that Nigeria was an adequate alternate forum. Id. at *16,18.

Several months later, the district court also granted Pfizer’s motion to dismiss the Adamu case. Adamu, 399 F.Supp.2d 495. It relied on its Abdullahi III decision to hold that the plaintiffs could not establish jurisdiction under the ATS. Id. at 501. The district court also incorporated the forum non conveniens analysis from Abdullahi III to find that Nigeria is an adequate forum. Id. at 504. Applying the public and private interest factors set forth in Gulf Oil Corp. v. Gilbert, 330 U.S. 501, 508-09, 67 S.Ct. 839, 91 L.Ed. 1055 (1947), superseded by statute on other grounds as *172recognized in Cowan v. Ford Motor Co., 713 F.2d 100, 103 (5th Cir. 1983), the court found that while public interest factors did not support either forum, private interest factors weighed in favor of dismissal. Adamu, 399 F.Supp.2d. at 505-06. The district court also dismissed the Adamu plaintiffs’ Connecticut law claims, concluding that, under Connecticut choice of law principles, the action was governed and barred by Nigerian law. Id. at 503.

The Abdullahi and Adamu plaintiffs appealed. Since then, a tectonic change has altered the relevant political landscape. In May 2007, the state of Kano brought criminal charges and civil claims against Pfizer, seeking over $2 billion in damages and restitution.4 Around the same time, the federal government of Nigeria sued Pfizer and several of its employees, seeking $7 billion in damages.5 None of these cases seek compensation for the subjects of the tests, who are the appellants before this Court. Pfizer then notified this Court that in light of these recent developments, which it believed required further consideration by the district court, it would not seek affirmance on the basis of forum non conveniens.

DISCUSSION

The district court dismissed both actions based on its determination that it lacked subject matter jurisdiction because plaintiffs failed to state claims under the ATS. We review dismissal on this ground de novo. Rweyemamu v. Cote, 520 F.3d 198, 201 (2d Cir. 2008). “To survive dismissal, the plaintiff[s] must provide the grounds upon which [their] claim rests through factual allegations sufficient ‘to raise a right to relief above the speculative level.’ ” ATSI Commc’ns, Inc. v. Shaar Fund, Ltd., 493 F.3d 87, 98 (2d Cir. 2007) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 127 5. Ct. 1955, 1965, 167 L.Ed.2d 929 (2007)).6

I. The Alien Tort Statute

The Alien Tort Statute, 28 U.S.C. § 1350, provides that “[t]he district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.” Included in the Judiciary Act of 1789, the statute provided jurisdiction in just two cases during the first 191 years after its enactment. See Taveras v. Taveraz, 477 F.3d 767, 771 (6th Cir. 2007). In the last thirty years, however, the ATS has functioned slightly more robustly, conferring jurisdiction over a limited category of claims.

We first extensively examined the ATS in Filartiga v. Pena-Irala, 630 F.2d 876 (2d Cir. 1980), where we held that conduct violating the law of nations is actionable under the ATS “only where the nations of the world have demonstrated that the wrong is of mutual, and not merely several, concern, by means of express international accords.” Id. at 888. Following *173Filartiga, we concluded that ATS claims may sometimes be brought against private actors, and not only state officials, see Kadic v. Karadzic, 70 F.3d 232, 239 (2d Cir. 1995), when the tortious activities violate norms of “universal concern” that are recognized to extend to the conduct of private parties — for example, slavery, genocide, and war crimes, id. at 240. This case involves allegations of both state and individual action. In Flores v. Southern Peru Copper Corp., 414 F.3d 233 (2d Cir. 2003), we clarified that “the law of nations” in the ATS context “refers to the body of law known as customary international law,” which “is discerned from myriad decisions made in numerous and varied international and domestic arenas” and “does not stem from any single, definitive, readily-identifiable source.” Id. at 247-48. These principles are rejected in their entirety by our dissenting colleague. In Flores, we concluded that ATS jurisdiction is limited to alleged violations of “those clear and unambiguous rules by which States universally abide, or to which they accede, out of a sense of legal obligation and mutual concern.” Id. at 252. Applying this standard, we held that the appellants’ claim that pollution from mining operations caused lung disease failed to state a violation of customary international law. We reasoned that the “right to life” and the “right to health” were insufficiently definite to constitute binding customary legal norms and that there was insufficient evidence to establish the existence of a narrower norm prohibiting intranational pollution. Id. at 254-55.

In 2004, the Supreme Court comprehensively addressed the ATS for the first time in Sosa v. Alvarez-Machain, 542 U.S. 692, 124 S.Ct. 2739, 159 L.Ed.2d 718 (2004). Justice Souter, writing for the majority, clarified that the ATS was enacted to create jurisdiction over “a relatively modest set of actions alleging violations of the law of nations” and with “the understanding that the common law would provide a cause of action.” Id. at 720, 723. The Supreme Court confirmed that federal courts retain a limited power to “adapt[] the law of nations to private rights” by recognizing “a narrow class of international norms” to be judicially enforceable through our residual common law discretion to create causes of action. Id. at 728-29. It cautioned, however, that courts must exercise this power with restraint and “the understanding that the door [to actionable violations] is still ajar subject to vigilant doorkeeping,” permitting only those claims that “rest on a norm of international character accepted by the civilized world and defined with a specificity comparable to the features of the 18th-century paradigms [the Supreme Court has] recognized.” Id. at 725, 729. These 18th-centu-ry paradigms consist of offenses against ambassadors, violations of the right to safe passage, and individual actions arising out of piracy. Id. at 724. The common theme among these offenses is that they contravened the law of nations, admitted of a judicial remedy, and simultaneously threatened serious consequences in international affairs. Id. at 715. Lower courts are required to gauge claims brought under the ATS against the current state of international law, but are permitted to recognize under federal common law only those private claims for violations of customary international law norms that reflect the same degree of “definite content and acceptance among civilized nations” as those reflected in the 18th-century paradigms. Id. at 732-33. The Supreme Court in Sosa also counseled that “the determination whether a norm is sufficiently definite to support a cause of action should (and, indeed, inevitably must) involve an element of judgment about the practical consequences of making that *174cause available to litigants” in federal courts. Id.

In this way Sosa set a “high bar to new private causes of action” alleging violations of customary international law. Id. at 727. A federal court can recognize one only if a plaintiff identifies the violation of a norm of customary international law that, as defined by the sources of such law that United States courts “have long, albeit cautiously, recognized,” id. at 733-34 (referencing The Paquete Habana, 175 U.S. 677, 700, 20 S.Ct. 290, 44 L.Ed. 320 (1900)), is sufficiently specific, universal, and obligatory to meet the standards established by Sosa. See Sosa, 542 U.S. at 732, 124 S.Ct. 2739 (citing with approval Tel-Oren v. Libyan Arab Republic, 726 F.2d 774, 781 (D.C.Cir. 1984) (Edwards, J., concurring), and In re Estate of Marcos, Human Rights Litig., 25 F.3d 1467, 1475 (9th Cir. 1994)). Applying these principles, the Supreme Court held that the plaintiff, a Mexican national who sued a fellow Mexican national under the ATS for allegedly aiding in his illegal abduction by agents of the U.S. Drug Enforcement Agency, had failed to allege the violation of a customary international law norm with the required precision. Sosa, 542 U.S. at 738, 124 S.Ct. 2739. The Supreme Court found that the practical consequences of recognizing a general and broad customary international law prohibition of arbitrary detention in a case involving “a single illegal detention of less than a day, followed by the transfer of custody to lawful authorities and a prompt arraignment” would be “breathtaking” and inappropriate. Id. at 736, 738, 124 S.Ct. 2739.

Since Sosa, this Court has reviewed three judgments dismissing claims under the ATS. In Khulumani v. Barclay National Bank, Ltd., 504 F.3d 254 (2d Cir. 2007) (per curiam), we held that the ATS conferred jurisdiction over multinational corporations that purportedly collaborated with the government of South Africa in maintaining apartheid because they aided and abetted violations of customary international law. Id. at 260. In Vietnam Ass’n for Victims of Agent Orange v. Dow Chemical Co., 517 F.3d 104 (2d Cir. 2008), we concluded that the ATS did not support a claim that the defendants violated international law by manufacturing and supplying Agent Orange and other herbicides used by the United States military during the Vietnam War. Id. at 123. We reasoned that the sources of law on which the appellants relied did not define a norm prohibiting the wartime use of Agent Orange that was both universal and sufficiently specific to satisfy the requirements of Sosa. Id. at 119-23. Similarly, in Mora v. People of the State of New York, 524 F.3d 183 (2d Cir. 2008), we held that the norm at issue — one that prohibits the detention of a foreign national without informing him of the requirement of consular notification and access under Article 36(1)(b)(3) of the Vienna Convention on Consular Relations — was insufficiently universal to support a claim under the ATS. Id. at 208-09.

Turning now to this appeal, and remaining mindful of our obligation to proceed cautiously and self-consciously in this area, we determine whether the norm alleged (1) is a norm of international character that States universally abide by, or accede to, out of a sense of legal obligation; (2) is defined with a specificity comparable to the 18th-century paradigms discussed in Sosa; and (3) is of mutual concern to States.

A The Prohibition of Nonconsensual Medical Experimentation on Humans

Appellants’ ATS claims are premised on the existence of a norm of custom*175ary international law prohibiting medical experimentation on non-consenting human subjects. To determine whether this prohibition constitutes a universally accepted norm of customary international law, we examine the current state of international law by consulting the sources identified by Article 38 of the Statute of the International Court of Justice (“ICJ Statute”), to which the United States and all members of the United Nations are parties. Flores, 414 F.3d at 250; see, e.g., United States v. Yousef, 327 F.3d 56, 100-01 (2d Cir. 2003). Article 38 identifies the authorities that provide “competent proof of the content of customary international law.” Flores, 414 F.3d at 251. These sources consist of:

(a) international conventions, whether general or particular, establishing rules expressly recognized by the contesting states;

(b) international custom, as evidence of a general practice accepted as law;

(c) the general principles of law recognized by civilized nations;

(d) ... judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law.

Statute of the International Court of Justice, art. 38(1), June 26,1945, 59 Stat. 1055, 1060, T.S. No. 993 [hereinafter ICJ Statute].

The appellants ground their claims in four sources of international law that categorically forbid medical experimentation on non-consenting human subjects: (1) the Nuremberg Code, which states as its first principle that “[t]he voluntary consent of the human subject is absolutely essential”; (2) the World Medical Association’s Declaration of Helsinki, which sets forth ethical principles to guide physicians world-wide and provides that human subjects should be volunteers and grant their informed consent to participate in research; (3) the guidelines authored by the Council for International Organizations of Medical Services (“CIOMS”), which require “the voluntary informed consent of [a] prospective subject”; and (4) Article 7 of the International Covenant on Civil and Political Rights (“ICCPR”), which provides that “no one shall be subjected without his free consent to medical or scientific experimentation.” 7

The district court found that “non-consensual medical experimentation violates the law of nations and, therefore, the laws of the United States” and cited the Nuremberg Code for support. Abdullahi III, 2005 WL 1870811, at *9. It then noted that “[w]hile federal courts have the authority to imply the existence of a private right of action for violations of jus cogens norms of international law, federal courts must consider whether there exist special factors counseling hesitation in the absence of affirmative action by Congress.” Id. (internal citations and quotation marks omitted). The district court then separately analyzed the four sources of international law that prohibit nonconsensual medical experimen*176tation on humans and the Universal Declaration of Human Rights. Id. at *11-13. It found that with the exception of the Nuremberg Code, these sources contain only aspirational or vague language lacking the specificity required for jurisdiction. Id. at *12-13. It also determined that because the United States did not ratify or adopt any of these authorities except the ICCPR, and because even the ICCPR is not self-executing, none of them create binding international legal obligations that are enforceable in federal court. Id. at *11-13. Finally, the district court concluded that the plaintiffs failed to provide a proper predicate for ATS jurisdiction because none of the sources independently authorizes a private cause of action and the inference of such a cause of action is a matter best left to Congress. Id. at *13-14.8

The district court’s approach misconstrued both the nature of customary international law and the scope of the inquiry required by Sosa. It mistakenly assumed that the question of whether a particular customary international law norm is sufficiently specific, universal, and obligatory to permit the recognition of a cause of action under the ATS is resolved essentially by looking at two things: whether each source of law referencing the norm is binding and whether each source expressly authorizes a cause of action to enforce the norm. But Sosa, as we have seen, requires a more fulsome and nuanced inquiry. Courts are obligated to examine how the specificity of the norm compares with 18th-century paradigms, whether the norm is accepted in the world community, and whether States universally abide by the norm out of a sense of mutual concern. By eschewing this inquiry, the district court did not engage the fact that norms of customary international law are “discerned from myriad decisions made in numerous and varied international and domestic arenas” and “[do] not stem from any single, definitive, readily-identifiable source.” Flores, 414 F.3d at 247-48.

The district court also inappropriately narrowed its inquiry in two respects. First, it focused its consideration on whether the norm identified by the plaintiffs is set forth in conventions to which the United States is a party, and if so, whether these treaties are self-executing or executed by federal legislation. While adoption of a self-executing treaty or the execution of a treaty that is not self-executing may provide the best evidence of a particular country’s custom or practice of recognizing a norm, see Flores, 414 F.3d at 257, the existence of a norm of customary international law is one determined, in part, by reference to the custom or practices of many States, and the broad acceptance of that norm by the international community. Agreements that are not self-executing or that have not been executed by federal legislation, including the ICCPR, are appropriately considered evidence of the current state of customary international law. See Khulumani, 504 F.3d at 284 (Katzmann, J., concurring) (noting that “[w]hether a treaty that embodies [a norm of customary international law] is self-executing is relevant to, but is not determinative of, [the] question” of whether the norm permits ATS jurisdiction). A formal treaty, moreover, is not the lone primary source of customary international law. The ICJ Statute permits, and Sosa encourages, among other things, that courts consider “international custom, as evidence of a general practice accepted *177as law.” ICJ Statute, supra, at art. 38(1); Sosa, 542 U.S. at 734, 124 S.Ct. 2739 (“[WJhere there is no treaty, and no controlling executive or legislative act or judicial decision, resort must be had to the customs and usages of civilized nations.”) (quoting The Paquete Habana, 175 U.S. at 700, 20 S.Ct. 290).

Second, the district court’s consideration of whether each source of law creates binding legal norms failed to credit the fact that even declarations of international norms that are not in and of themselves binding may, with time and in conjunction with state practice, provide evidence that a norm has developed the specificity, universality, and obligatory nature required for ATS jurisdiction. See Filartiga, 630 F.2d at 883 (“[A non-binding] Declaration creates an expectation of adherence, and insofar as the expectation is gradually justified by State practice, a declaration may by custom become recognized as laying down rules binding upon the States.”) (internal quotation marks omitted). The district court should have considered a greater range of evidence and weighed differently the probative value of the sources on which the appellants relied.

In sum, it was inappropriate for the district court to forego a more extensive examination of whether treaties, international agreements, or State practice have ripened the prohibition of nonconsensual medical experimentation on human subjects into a customary international law norm that is sufficiently (i) universal and obligatory, (ii) specific and definable, and (iii) of mutual concern, to permit courts to infer a cause of action under the ATS. See Sosa, 542 U.S. at 732-35, 124 S.Ct. 2739. We now proceed with such an examination.

i. Universality

The appellants must allege the violation of a norm of customary international law to which States universally subscribe. See Sosa, 542 U.S. at 732, 124 S.Ct. 2739; Vietnam Ass’n for Victims of Agent Orange, 517 F.3d at 117. The prohibition on nonconsensual medical experimentation on human beings meets this standard because, among other reasons, it is specific, focused and accepted by nations around the world without significant exception.

The evolution of the prohibition into a norm of customary international law began with the war crimes trials at Nuremberg. The United States, the Soviet Union, the United Kingdom and France “acting in the interest of all the United Nations,” established the International Military Tribunal (“IMT”) through entry into the London Agreement of August 8, 1945. M. Cheriff Bassiouni et al., An Appraisal of Human Experimentation in International Law and Practice: The Need for International Regulation of Human Experimentation, 72 J.Crim. L. & Criminology 1597, 1640 & n. 220 (1981) (internal quotation marks omitted). Annexed to the London Agreement was the London Charter, which served as the IMT’s Constitution. See Agreement for the Prosecution and Punishment of the Major War Criminals of the European Axis Powers, with annexed Charter of the International Military Tribunal art. 2, Aug. 8, 1945, 59 Stat. 1544, 82 U.N.T.S. 279. According to the Charter, the IMT had the “power to try and punish persons who, acting in the interests of the European Axis countries, whether as individuals or as members of organizations, committed,” among other offenses, war crimes and crimes against humanity. Id. at art. 6.

The IMT tried 22 “major” Nazi war criminals leaving “lower-level” war criminals, including “[l]eading physicians ... and leading German industrialists,” to be tried in subsequent trials by U.S. military tribunals acting “under the aegis of the *178IMT.” United States Holocaust Memorial Museum, War Crimes Trials, Holocaust Encylopedia (2008), http://www.ushmm. org/wlc/article.php?lang=en&ModuleId= 10005140. The law that authorized the creation of the U.S. military tribunals, Control Council Law No. 10, was enacted in 1945 by the Allied Control Council, see id., an authority through which the London Agreement signatories exerted joint-control over Germany, see Encyclopedia Britannica, Germany, Encyclopedia Britannica Online (2009), http://search.eb.com/ eb/article-58214. Control Council Law No. 10 stated that its purpose was to “give effect to the terms of ... the London Agreement ... and the [London] Charter,” and “to establish a uniform legal basis in Germany for the prosecution of war criminals.” Allied Control Council No. 10, preamble, (Dec. 20, 1945), http://avalon. Iaw.yale.edu/imt/imtl0.asp. Law No. 10 expressly incorporated the London Agreement, identifying it as an “integral part[ ] of this Law.” Id. at art. I. Law No. 10 also authorized military tribunals of the occupying powers to prosecute individuals for the same crimes over which the IMT had jurisdiction, including war crimes and crimes against humanity, see id. at arts. II — III, and made military tribunal prosecutions subject to the IMT’s right of first refusal, see id. at art. III. Consequently, the U.S. military tribunals effectively operated as extensions of the IMT, see Telford Taylor, Final Report to the Secretary of the Army on the Nuernberg War Crimes Trials Under Control Council Law No. 107, 107 (1949) [hereinafter Report on Nuernberg War Crimes Trials], available at http://www.loc.gov/rr/frd/Military_Law/ pdf/NT_final-report.pdf (explaining that “the trials under Law No. 10 were to be a means of carrying out such 'declarations of criminality’ ... as the International Military Tribunal might make” and that “[t]he first [IMT] trial and the 12 following [military tribunal] trials ... form a single sequence based on common principles”), and Control Council Law No. 10 served to implement the commitments undertaken in the London Agreement, see id. at 7 (noting that “the two documents supplemented each other” and “[m]ajor criminals not tried under the one could be tried under the other”).

In August 1947, Military Tribunal 1, staffed by American judges and prosecutors and conducted under American procedural rules, see George J. Annas, The Nuremberg Code in U.S. Courts: Ethics versus Expediency, in The Nazi Doctors and the Nuremberg Code 201, 201 (George J. Annas & Michael A. Grodin eds., 1992), promulgated the Nuremberg Code as part of the tribunal’s final judgment against fifteen doctors who were found guilty of war crimes and crimes against humanity for conducting medical experiments without the subjects’ consent, Brandt, 2 Nuremberg Trials, at 181-82. Among the nonconsensual experiments that the tribunal cited as a basis for their convictions were the testing of drugs for immunization against malaria, epidemic jaundice, typhus, smallpox and cholera. Id. at 175-178. Seven of the convicted doctors were sentenced to death and the remaining eight were sentenced to varying terms of imprisonment. Id. at 298-300. The tribunal emphasized that

[i]n every single instance appearing in the record, subjects were used who did not consent to the experiments; indeed, as to some of the experiments, it is not even contended by the defendants that the subjects occupied the status of volunteers.

Id. at 183. The judgment concluded that “[manifestly human experiments under such conditions are contrary to the principles of the law of nations as they result from usages established among civilized *179peoples, from the laws of humanity, and from the dictates of public conscience.” Id. (emphasis added and internal quotation marks omitted). The Code created as part of the tribunal’s judgment therefore emphasized as its first principle that “[t]he voluntary consent of the human subject is absolutely essential.” Id. at 181.

The American tribunal’s conclusion that action that contravened the Code’s first principle constituted a crime against humanity is a lucid indication of the international legal significance of the prohibition on nonconsensual medical experimentation. As Justices of the Supreme Court have recognized, “[t]he medical trials at Nuremberg in 1947 deeply impressed upon the world that experimentation with unknowing human subjects is morally and legally unacceptable.” United States v. Stanley, 483 U.S. 669, 687, 107 S.Ct. 3054, 97 L.Ed.2d 550 (1987) (Brennan, J., concurring in part and dissenting in part) (emphasis added); see also id. at 709-10, 107 S.Ct. 3054 (O’Connor, J., concurring in part and dissenting in part).

Moreover, both the legal principles articulated in the trials’ authorizing documents and their application in judgments at Nuremberg occupy a position of special importance in the development of bedrock norms of international law. United States courts examining the Nuremberg judgments have recognized that “[tjhe universal and fundamental rights of human beings identified by Nuremberg — rights against genocide, enslavement, and other inhumane acts ... — are the direct ancestors of the universal and fundamental norms recognized as jus cogens,” from which no derogation is permitted, irrespective of the consent or practice of a given State. Siderman de Blake v. Republic of Arg., 965 F.2d 699, 715 (9th Cir. 1992) (cited in Sampson v. F.R.G., 250 F.3d 1145, 1150 (7th Cir. 2001)). As Telford Taylor, who first served as an assistant to Justice Robert Jackson during his time as Chief Prosecutor for the IMT and then became Chief of Counsel for War Crimes on the Nuremberg trials held under the authority of Control Council Law No. 10, explained, “Nuernberg was based on enduring [legal] principles and not on temporary political expedients, and this fundamental point is apparent from the reaffirmation of the Nuernberg principles in Control Council Law No. 10, and their application and refinement in the 12 judgments rendered under that law during the 3-year period, 1947 to 1949.” Taylor, Report on Nuernberg War Crimes Trials, at 107 (emphasis added).

Consistent with this view, the Code’s first principle has endured: “[Significant world opinion has not come to the defense of the nature or manner in which the experiments were conducted in the Nazi concentration camps.” Bassiouni et al., supra, at 1641. Rather, since Nuremberg, states throughout the world have shown through international accords and domestic law-making that they consider the prohibition on nonconsensual medical experimentation identified at Nuremberg as a norm of customary international law.9

*180In 1955, the draft International Covenants on Human Rights was revised to add a second sentence to its prohibition of torture and cruel, inhuman or degrading treatment or punishment. The addition provided that “[i]n particular, no one shall be subjected without his free consent to medical or scientific experimentation involving risk, where such is not required by his state of physical or mental health.” Annotations on the text of the draft International Covenants on Human Rights, at 31, U.N. GAOR, 10th Sess., Annexes, agenda item 28(11), U.N. Doc. A/2929 (July 1, 1955). The clause was later revised to offer the simpler and sweeping prohibition that “no one shall be subjected without his free consent to medical or scientific experimentation.” ICCPR, supra, at art. 7. This prohibition became part of Article 7 of the ICCPR, which entered into force in 1976, and is legally binding on the more than 160 States-Parties that have ratified the convention without reservation to the provision.10 By its terms this prohibition is not limited to state actors; rather, it guarantees individuals the right to be free from nonconsensual medical experimentation by any entity—state actors, private actors, or state and private actors behaving in concert.

Its status as a norm that states conceive as legally binding — and therefore part of customary international law — is confirmed by Article 2 of the accord, which requires that “[e]ach State Party ... undertake[ ] to respect and to ensure to all individuals within its territory and subject to its jurisdiction the rights recognized in the present Covenant.” ICCPR art. 2(1). The international community’s recognition in the ICCPR of its obligation to protect humans against nonconsensual medical experimentation, regardless of the source of the action, is powerful evidence of the prohibition’s place in customary international law.

It is clear that, as the court mentioned in Sosa, the Universal Declaration of Human Rights and the ICCPR themselves could not establish the relevant, applicable rule of international law in that case. Sosa, 542 U.S. at 754, 124 S.Ct. 2739. Nonetheless, the ICCPR, when viewed as a reaffirmation of the norm as articulated in the Nuremberg Code, is potent authority for the universal acceptance of the prohibition on nonconsensual medical experimentation. As we discuss below, see infra pp. 181-83, the fact that the prohibition on medical experimentation on humans without consent has been consciously embedded by Congress in our law and reaffirmed *181on numerous occasions by the FDA demonstrates that the United States government views the norm as the source of a binding legal obligation even though the United States has not ratified the ICCPR in full.11

In 1964, the World Medical Association adopted the Declaration of Helsinki, which enunciated standards for obtaining informed consent from human subjects. It provided that in clinical research combined with professional care, “[i]f at all possible, consistent with patient psychology, the doctor should obtain the patient’s freely given consent after the patient has been given a full explanation,” and that non-therapeutic clinical research on a person “cannot be undertaken without his free consent, after he has been fully informed.” World Med. Ass’n, Declaration of Helsinki: Code of Ethics of the World Medical Association, art. III(3a), G.A. Res. (1964), http://www.pubmedcentral.nih.gov/ picrender.fcgi?artid= 1816102 & blob-type =pdf. The Declaration has since been amended five times. The informed consent provision now provides that “subjects must be volunteers and informed participants in the research project.” Declaration of Helsinki, supra, at art. 20. The Declaration also requires that “[i]n any research on human beings, each potential subject must be adequately informed of the aims, methods, ... anticipated benefits and potential risks of the study, and the discomfort it may entail” and that researchers “obtain the subject’s freely-given informed consent, preferably in writing.” Id. at art. 22.

Although the Declaration itself is nonbinding, since the 1960s, it has spurred States to regulate human experimentation, often by incorporating its informed consent requirement into domestic laws or regulations. See Delon Human & Sev S. Fluss, The World Medical Association’s Declaration of Helsinki: Historical and Contemporary Perspectives, 8-11 (July 24, 2001) (fifth draft), http://www.wma.neUe/ ethicsuniUpdf/draft_historical_ contemporary_perspectives.pdf (describing legal and regulatory developments in Australia, Belgium, Brazil, China, Israel, Japan, New Zealand, Norway, Switzerland, and the United States following the Declaration of Helsinki). Currently, the laws and regulations of at least eighty-four countries, including the United States, require the informed consent of human subjects in medical research.12 That this conduct has been the subject of domestic legislation is not, of course, in and of itself proof of a norm. See Flores, 414 F.3d at 249. However, the incorporation of this norm into the laws of this country and this host of others is a powerful indication of the international acceptance of this norm as a binding legal obligation, where, as here, states have shown that the norm is of mutual concern by including it in a variety of international accords.

*182The history of the norm in United States law demonstrates that it has been firmly embedded for more than 45 years and-— except for our dissenting colleague — its validity has never been seriously questioned by any court. Congress mandated patient-subject consent in drug research in 1962. Bassiouni et al., supra, at 1624 (citing 21 U.S.C. § 355(i) (1976)). In response, the FDA promulgated its first regulations requiring the informed consent of human subjects. Tellingly, the sources on which our government relied in outlawing non-consensual human medical experimentation were the Nuremberg Code and the Declaration of Helsinki, which suggests the government conceived of these sources’ articulation of the norm as a binding legal obligation. Bassiouni et al., supra, at 1625-26 (citing 21 C.F.R. § 310.102(h) (1980)).13 Today, FDA regulations require informed consent to U.S. investigators’ research, whether conducted domestically or in a foreign country, used to support applications for the approval of new drugs. See 21 C.F.R. §§ 50.20, 50.23-.25, 50.27, 312.20, 312.120 (2008); 45 C.F.R. §§ 46.111, 46.116 — .117 (2008).

The importance that the United States government attributes to this norm is demonstrated by its willingness to use domestic law to coerce compliance with the norm throughout the world. United States law requires that, as a predicate to FDA approval of any new drug, both American and foreign sponsors of drug research involving clinical trials, whether conducted here or abroad, procure informed consent from human subjects. 21 C.F.R. §§ 312.20, 312.120 (2008); see also Dep’t of Health & Human Servs., Office of Inspector Gen., The Globalization of Clinical Trials 5 (2001), http://www.oig.hhs.gov/ oei/reports/oei-01-00-00190.pdf. Sponsors conducting research under an Investigational New Drug Application (“IND”) are obligated to adhere to FDA regulations, which require informed consent. 21 C.F.R. § 312.20 (2008); The Globalization of Clinical Trials, supra, at 5. Prior to April 2008, sponsors conducting research under non-IND guidelines were obligated to adhere to the ethical principles of the 1989 version of the Declaration of Helsinki or the host country’s regulations, whichever offered greater protection to the human subject. 21 C.F.R. § 312.120 (2007); The Globalization of Clinical Trials, supra, at 5. The April 2008 revisions to the non-IND guidelines reaffirmed the informed consent requirement. Human Subject Protection: Foreign Clinical Studies Not Conducted Under an Investigational New Drug Application, 73 Fed.Reg. 22,800, 22,801, 22,803, *18322,804, 22,816 (Apr. 28, 2008) (codified at 21 C.F.R. pt. 312). Foreign clinical studies not conducted under an IND must now comply with the Good Clinical Practice guidelines (“GCP”) promulgated by the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, 62 Fed.Reg. 25,692 (May 9, 1997), which require informed consent to medical experimentation. 21 C.F.R. § 312.120 (2008).

Additional international law sources support the norm’s status as customary international law. The European Union embraced the norm prohibiting nonconsensual medical experimentation through a 2001 Directive passed by the European Parliament and the Council of the European Union. The Directive accepted the informed consent principles of the 1996 version of the Declaration of Helsinki. Council Directive 2001/20/EC, preamble (2), 2001 O.J. (L 121) 37(EC) [hereinafter 2001 Clinical Trial Directive]. It also required member States to adopt rules protecting individuals incapable of giving informed consent and permitting clinical trials only where “the trial subject or, when the person is not able to give informed consent, his legal representative has given his written consent after being informed of the nature, significance, implications and risks of the clinical trial.” Id. at art. (1), (2)(d). The Directive further required all member States to implement by 2004 domestic laws, regulations, and administrative provisions to comply with its informed consent requirements. Id. at art. 22(1).

Since 1997, thirty-four member States of the Council of Europe have also signed the Convention on Human Rights and Biomedicine, a binding convention and a source of customary international law. Convention for the Protection of Human Rights and Dignity of the Human Being with regard to the Application of Biology and Medicine: Convention on Human Rights and Biomedicine, art. 5, 15-16, opened for signature Apr. 4, 1997, E.T.S. No. 164, http:// conventions.coe.int/Treaty/en/Treaties/ html/164.htm [hereinafter Convention on Human Rights and Biomedicine]; Convention on Human Rights and Biomedicine, Chart of Signatures and Ratifications as of Aug. 8, 2008, http://conventions.coe.inV Treaty/Commun/ChercheSig.asp? NT=164 & CM=8 & DF=8/8/2008 & CL=ENG. It provides that an “intervention in the health field may only be carried out after the person concerned has given free and informed consent to it” and that the informed consent of human subjects is required for their involvement in medical research. Convention on Human Rights and Biomedicine, supra, at art. 5.14 In 2005, the General Conference of the United Nations Educational, Scientific and Cultural Organization (UNESCO) adopted the Universal Declaration on Bioethics and Human Rights, which requires “the prior, free, express and informed consent of the person concerned” for research-oriented treatments. Universal Declaration on Bioethics and Human Rights, UNESCO Gen. Conf. Res., at art. 6, 33rd Sess., 33 C/Resolution 36, (Oct. 19, 2005).

This history illustrates that from its origins with the trial of the Nazi doctors at Nuremburg through its evolution in international conventions, agreements, declarations, and domestic laws and regulations, the norm prohibiting nonconsensual medical experimentation on human subjects has become firmly embedded and has secured *184universal acceptance in the community of nations. Unlike our dissenting colleague’s customary international law analysis, which essentially rests on the mistaken assumption that ratified international treaties are the only valid sources of customary international law for ATS purposes, see Dissent at 200-02, we reach this conclusion as a result of our review of the multiplicity of sources — including international conventions, whether general or particular, and international custom as identified through international agreements, declarations and a consistent pattern of action by national law-making authorities — that our precedent requires us to examine for the purpose of determining the existence of a norm of customary international law. Our dissenting colleague’s reasoning fails to engage the incompatibility of nonconsensual human testing with key sources of customary international law identified in Article 38 of the ICJ’s statute, most importantly international custom, as evidence of a general practice accepted as law, as well as the general principles of law recognized by civilized nations. See supra pp. 174-75.

ii Specificity

Sosa requires that we recognize causes of action only to enforce those customary international law norms that are no “less definite [in] content ... than the historical paradigms familiar when [the ATS] was enacted.” Sosa, 542 U.S. at 732, 124 S.Ct. 2739. The norm prohibiting nonconsensual medical experimentation on human subjects meets this requirement. In United States v. Smith, 18 U.S. (5 Wheat) 153, 159-61, 5 L.Ed. 57 (1820), Justice Story found that “whatever may be the diversity of definitions, ... all writers concur, in holding, that robbery or forcible depredations upon the sea ... is piracy.” Id. at 161. We have little trouble concluding that a norm forbidding nonconsensual human medical experimentation is every bit as concrete — indeed even more so — than the norm prohibiting piracy that Story describes, or interference with the right of safe conducts and the rights of ambassadors, which together are the paradigmatic norms identified in Sosa. Id. at 724, 124 S.Ct. 2739. The Nuremberg Code, Article 7 of the ICCPR, the Declaration of Helsinki, the Convention on Human Rights and Biomedicine, the Universal Declaration on Bioethics and Human Rights, the 2001 Clinical Trial Directive, and the domestic laws of at least eighty-four States all uniformly and unmistakably prohibit medical experiments on human beings without their consent, thereby providing concrete content for the norm.15 The appellants allege that Pfizer knowingly and purposefully conducted such experiments on a large scale. Whatever uncertainty may exist at the margin is irrelevant here because appellants allege a complete failure on the part of Pfizer and the Nigerian government to inform appellants of the existence of the Trovan experiments. These allegations, if true, implicate Pfizer and the Nigerian government in conduct that is at the core of any reasonable iteration of the prohibition against involuntary *185medical experimentation. While the prohibition in question applies to the testing of drugs without the consent of human subjects on the scale Pfizer allegedly conducted, we do not suggest that it would extend to instances of routine or isolated failures by medical professionals to obtain informed consent, such as those arising from simple negligence. The allegations in the complaints involve anything but a doctor’s routine or erroneous failure to obtain such consent from his patient.

Hi. Mutual Concern

Customary international law proscribes only transgressions that are of “mutual” concern to States — “those involving States’ actions performed ... towards or with regard to the other.” Flores, 414 F.3d at 249 (differentiating matters of “mutual” concern from those of “several” concern, in which “States are separately and independently interested”). Conduct that States have prohibited through domestic legislation is also actionable under the ATS as a violation of customary international law when nations of the world have demonstrated “by means of express international accords” that the wrong is of mutual concern. Filartiga, 630 F.2d at 888. An important, but not exclusive, component of this test is a showing that the conduct in question is “capable of impairing international peace and security.” Flores, 414 F.3d at 249. Appellants have made both of these showings.

As we have seen, States throughout the world have entered into two express and binding international agreements prohibiting nonconsensual medical experimentation: the ICCPR and the Convention on Human Rights and Biomedicine. The entry of over 160 States into these agreements and the European Union’s passage of the 2001 Clinical Trial Directive demonstrates that States have not only acted independently to outlaw large-scale, non-consensual drug testing on humans, but they have also acted in concert to do so. In other words, acting out of a sense of mutual concern, “the nations [of the world] have made it their business, both through international accords and unilateral action,” to demonstrate their intention to eliminate conduct of the type alleged in the complaints. Filartiga, 630 F.2d at 889.

The administration of drug trials without informed consent on the scale alleged in the complaints poses a real threat to international peace and security. Over the last two decades, pharmaceutical companies in industrialized countries have looked to poorer, developing countries as sites for the medical research essential to the development of new drugs. See James V. Lavery, Putting International Research Ethics Guidelines to Work for the Benefit of Developing Countries, 4 Yale J. Health Pol’y L. & Ethics 319, 320-21 (2004); The Globalization of Clinical Trials, supra, at 8.16 Pharmaceutical companies recognize the potential benefits of drug trials to poor nations and have sought to promote access to medicines and health care in under-served populations through philanthropy and partnership with governments and NGOs. See, e.g., PhRMA, Press Releases: Worldwide Pharmaceutical Industry Launches Global Health Progress Initiative to Expand Efforts to Improve Health in Developing Countries (April 16, 2008), http://www.phrma.org/news.jroom/press_ releases/global_health_progress_initiative_ launched_to_improve_health_in_deve loping_countries/ (describing initiative by worldwide pharmaceutical industry to “fur*186ther access to medicines; build capacity of health workers in developing nations; advocate for global action to address health challenges; and continue R & D to develop new tools to fight diseases that plague the developing world”); PhRMA, Profile2008: Pharmaceutical Industry 42 (2008), http:// www.phrma.org/files/2008% 20Profile.pdf (describing contributions by American pharmaceutical companies to the promotion of global access to medicines and health care). This trend offers the possibility of enormous health benefits for the world community. Life-saving drugs can potentially be developed more quickly and cheaply, and developing countries may be given access to cutting edge medicines and treatments to assist underresourced and understaffed public health systems, which grapple with life-threatening diseases afflicting their populations.17

The success of these efforts promises to play a major role in reducing the cross-border spread of contagious diseases, which is a significant threat to international peace and stability. The administration of drug trials without informed consent on the scale alleged in the complaints directly threatens these efforts because such conduct fosters distrust and resistance to international drug trials, cutting edge medical innovation, and critical international public health initiatives in which pharmaceutical companies play a key role. This case itself supplies an exceptionally good illustration of why this is so. The Associated Press reported that the Trovan trials in Kano apparently engendered such distrust in the local population that it was a factor contributing to an eleven month-long, local boycott of a polio vaccination campaign in 2004, which impeded international and national efforts to vaccinate the population against a polio outbreak with catastrophic results.18 According to the World Health Organization, polio originating in Nigeria triggered a major international outbreak of the disease between 2003 and 2006, causing it to spread across west, central, and the Horn of Africa and *187the Middle East, and to re-infect twenty previously polio-free countries.19

The administration of drug trials without informed consent also poses threats to national security by impairing our relations with other countries. Seven of the world’s twelve largest pharmaceutical manufacturers — a group that includes Pfizer — are American companies. Global 500, Fortune, July 21, 2008, http://money.cnn. com/magazines/fortune/global500/2008/ industries/21/index.html. Consequently, American companies are likely to be sponsors of medical experiments on human subjects abroad.20 As this case illustrates, the failure to secure consent for human experimentation has the potential to generate substantial anti-American animus and hostility. Unsurprisingly, as noted above, see supra pp. 201-02, our government actively attempts to prevent this practice in foreign countries. For example, federal law requires that data generated from testing on human subjects abroad that is used to seek regulatory approval for a given drug must, at minimum, be the result of testing conducted consistent with the requirements of informed consent. Consequently, the U.S. government denies access to the U.S. market for any new drug unless the drug’s research data is generated in a manner consistent with the customary international law norm prohibiting drug trials on human subjects without informed consent.

For these reasons, we hold that the appellants have pled facts sufficient to state a cause of action under the ATS for a violation of the norm of customary international law prohibiting medical experimentation on human subjects without their consent. In such an instance, ATS jurisdiction exists over plaintiffs’ claims. The district court determined that the norm existed, but concluded that because no single source recognizing the norm was legally binding on the United States and created a private cause of action, it could not infer such a right under the ATS. Presumably, on this basis, it simultaneously held that there was no subject matter jurisdiction over plaintiffs’ claims. Under Sosa, this approach was not correct. Sosa makes clear that the critical inquiry is whether the variety of sources that we are required to consult establishes a customary international law norm that is sufficiently specific, universally accepted, and obligatory for courts to recognize a cause of action to enforce the norm. Nothing in Sosa suggests that this inquiry can be halted if some of the sources of international law giving rise to the norm are found not to be binding or not to explicitly authorize a cause of action.

We believe that the issues raised by this appeal regarding customary international *188law are framed by our analysis and by that of our dissenting colleague. He contends that our analysis is created from “whole cloth.” Dissent at 191. We believe that his approach to customary international law is unselfconsciously reactionary and static. The approach does not accommodate itself to the normative world that, by their commitments and conduct over the past fifty years, states — including our own — have shown they believe to exist.

B. State Action

A private individual will be held liable under the ATS if he “acted in concert with” the state, i.e., “under color of law.” Kadic, 70 F.3d at 245. In making this determination, courts look to the standards developed for finding state action in claims brought under 42 U.S.C. § 1983. Id. Under § 1983, state action may be found when “there is such a ‘close nexus between the State and the challenged action’ that seemingly private behavior ‘may be fairly treated as that of the State itself.’ ” Brentwood Acad. v. Tenn. Secondary Sch. Athletic Ass’n, 531 U.S. 288, 295, 121 S.Ct. 924, 148 L.Ed.2d 807 (2001) (quoting Jackson v. Metropolitan Edison Co., 419 U.S. 345, 351, 95 S.Ct. 449, 42 L.Ed.2d 477 (1974)). That nexus may exist “where a private actor has operated as a willful participant in joint activity with the State or its agents,” Gorman-Bakos v. Cornell Coop. Extension of Schenectady County, 252 F.3d 545, 551-52 (2d Cir. 2001) (quoting Loce v. Time Warner Entertainment Advance/Newhouse Partnership, 191 F.3d 256, 266 (2d Cir. 1999)), or “acts together with state officials or with significant state aid,” Kadic, 70 F.3d at 245. Pfizer meets this test.

The Appellants have alleged that the Nigerian government was involved in all stages of the Kano test and participated in the conduct that violated international law. They allege that the Nigerian government provided a letter of request to the FDA to authorize the export of Trovan, arranged for Pfizer’s accommodations in Kano, and facilitated the nonconsensual testing in Nigeria’s IDH in Kano. Despite overcrowding due to concurrent epidemics, the Nigerian government extended the exclusive use of two hospital wards to Pfizer, providing Pfizer with control over scarce public resources and the use of the hospital’s staff and facilities to conduct the Kano test, to the exclusion of MSF.

The unlawful conduct is alleged to have occurred in a Nigerian facility with the assistance of the Nigerian government and government officials and/or employees from the IDH and Aminu Kano Teaching Hospital. Pfizer’s research team in Kano was comprised of three American physicians, Dr. Abdulhamid Isa Dutse (a physician in the Aminu Kano Teaching Hospital), and three other Nigerian doctors. The American and Nigerian members of Pfizer’s team allegedly jointly administered the Kano test. Finally, in addition to assisting with the Kano test, Nigerian officials are alleged to have conspired to cover up the violations by silencing Nigerian physicians critical of the test and by back-dating an “approval letter” that the FDA and international protocol required to be provided prior to conducting the medical experiment. In addition to these allegations, the Adamu plaintiffs explicitly allege that the Nigerian government “was intimately involved and contributed, aided, assisted and facilitated Pfizer’s efforts to conduct the Trovan test,” “acted in concert with Pfizer,” and, according to a Nigerian physician involved in the Trovan experimentation, appeared to “back[ ]” the testing. At the pleading stage, these contentions meet the state action test because they adequately allege that the violations occurred as the result of concerted action *189between Pfizer and the Nigerian government.

II. Forum Non Conveniens

As an alternative to dismissal for failure to state a claim under the ATS, the district court dismissed the actions on the ground of forum non conveniens. Appellants raised this issue on appeal. Ordinarily, we review a forum non conveniens dismissal for abuse of discretion. Norex Petroleum Ltd. v. Access Indus., Inc., 416 F.3d 146, 153 (2d Cir. 2005). Since filing this appeal, however, Pfizer has notified the Court that in light of recent developments, in particular the initiation of proceedings by the federal government of Nigeria and the state of Kano against Pfizer and certain of its employees, it would not seek affirmance of the judgment on the basis of forum non conveniens. The appellants agreed and also requested that the issue be remanded. We accede to this request.

Although we are not now called upon definitively to review the district court’s application of forum non conveniens, in view of the frequency with which this issue has arisen and remained unsettled in this case, we offer additional guidance to assist the parties and the district court. The three-step analysis set forth in Iragorri v. United Techs. Corp., 274 F.3d 65, 71-75 (2d Cir. 2001) (en banc), applies. In this litigation, the second step of the analysis, which requires the district court to consider the adequacy of the alternative forum, is pivotal. Dismissal is not appropriate if an adequate and presently available alternative forum does not exist. Norex, 416 F.3d at 159. A forum in which defendants are amenable to service of process and which permits litigation of the dispute is generally adequate. Id. at 157. Such a forum may nevertheless be inadequate if it does not permit the reasonably prompt adjudication of a dispute, if the forum is not presently available, or if the forum provides a remedy so clearly unsatisfactory or inadequate that it is tantamount to no remedy at all. Piper Aircraft Co. v. Reyno, 454 U.S. 235, 254-55 & n. 22, 102 S.Ct. 252, 70 L.Ed.2d 419 (1981); USHA (India), Ltd. v. Honeywell Int’l, Inc., 421 F.3d 129, 136 (2d Cir. 2005); Norex, 416 F.3d at 160.

The defendant bears the burden of establishing that a presently available and adequate alternative forum exists, and that the balance of private and public interest factors tilts heavily in favor of the alternative forum. USHA (India), Ltd., 421 F.3d at 135; PT United Can Co. v. Crown Cork & Seal Co., Inc., 138 F.3d 65, 74 (2d Cir. 1998). Absent a showing of inadequacy by a plaintiff, “considerations of comity preclude a court from adversely judging the quality of a foreign justice system.” PT United Can Co., 138 F.3d at 73. Accordingly, while the plaintiff bears the initial burden of producing evidence of corruption, delay or lack of due process in the foreign forum, the defendant bears the ultimate burden of persuasion as to the adequacy of the forum. See, e.g., Norex, 416 F.3d at 159-160.

When the district court granted Pfizer’s motion, it identified the pivotal issue as whether the plaintiffs produced sufficient evidence to show that Nigeria is an inadequate alternative forum. Abdullahi III, 2005 WL 1870811, at *15. Having found that they had not, it concluded that Nigeria was an adequate forum. Id. at *16-18. In so doing, the district court omitted an analysis of whether Pfizer discharged its burden of persuading the court as to the adequacy and present availability of the Nigerian forum and improperly placed on plaintiffs the burden of proving that the alternative forum is inadequate. Cf. DiRienzo v. Philip Servs. Corp., 294 F.3d 21, 30 (2d Cir. 2002) (holding that it is error *190not “to hold defendants to their burden of proof’ of the Gilbert factors). On remand, the district court will have an opportunity to reassess this issue, as well as the relationship between Fed.R.Civ.P. 44.1 and the Federal Rules of Evidence.

III. Choice of Law

The district court dismissed the Adamu plaintiffs’ claims under the Connecticut Unfair Trade Practices Act and the Connecticut Products Liability Act on the ground that Connecticut choice of law principles applied and called for the application of Nigerian law. Adamu, 399 F.Supp.2d at 501-03. “We review the district court’s choice of law de novo.” Fin. One Pub. Co. Ltd. v. Lehman Bros. Special Fin., Inc., 414 F.3d 325, 331 (2d Cir. 2005).

The district court correctly determined that Connecticut choice-of-law rules applied because it was obligated to apply the state law that would have been applicable if the ease had not been transferred from Connecticut to New York. See Van Dusen v. Barrack, 376 U.S. 612, 639, 84 S.Ct. 805, 11 L.Ed.2d 945 (1964). Under Connecticut law, lex loci delicti, “the doctrine that the substantive rights and obligations arising out of a tort controversy are determined by the law of the place of injury,” typically applies. O’Connor v. O’Connor, 201 Conn. 632, 637, 519 A.2d 13 (1986). Lex loci delicti would require the application of Nigerian law because the Adamu plaintiffs’ injuries are alleged to have occurred there. Connecticut, however, has conspicuously retreated from a rigid application of the doctrine. The Connecticut Supreme Court held that lex loci delicti does not apply to a tort claim when doing so would undermine expectations of the parties or an important state policy, produce an arbitrary and irrational result, or where “reason and justice” counsel for the application of a different principle. Id. at 637, 648, 650, 519 A.2d 13. In such cases, Connecticut courts are required to apply the “most significant relationship” analysis set forth in the Restatement (Second) of Conflict of Laws §§ 6 & 145 (1971) [hereinafter Restatement (Second)]. O’Connor, 201 Conn, at 649-50, 519 A.2d 13.

Section 145(1) of the Restatement provides that “[t]he rights and liabilities of the parties with respect to an issue in tort are determined by the local law of the state which, with respect to that issue, has the most significant relationship to the occurrence and the parties under the principles stated in § 6.” Restatement (Second) § 145(1). Section 6(2), in turn, provides that where a state is not guided by a statutory directive on choice of law,

the factors relevant to the choice of the applicable rule of law include

(a) the needs of the interstate and international systems,

(b) the relevant policies of the forum,

(c) the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue,

(d) the protection of justified expectations,

(e) the basic policies underlying the particular field of law,

(f) certainty, predictability and uniformity of result, and

(g) ease in the determination and application of the law to be applied.

Restatement (Second) § 6(2). The Connecticut Supreme Court has determined that Section 145(2) provides courts with guidance regarding the evaluation of the policy choices set out in Sections 145(1) and 6(2). O’Connor, 201 Conn. at 652, 519 A.2d 13. Section 145(2) assists with the *191application of the principles of Section 6 to tort cases by calling for consideration of:

(a) the place where the injury occurred,

(b) the place where the conduct causing the injury occurred,

(c) the domicil, residence, nationality, place of incorporation and place of business of the parties, and

(d) the place where the relationship, if any, between the parties is centered.

Restatement (Second) § 145(2). These factors are “to be evaluated according to their relative importance with respect to the particular issue.” Id.

The district court correctly decided to apply Sections 6 and 145 of the Restatement rather than lex loci delicti. It applied the factors in Section 145(2) to determine whether Connecticut or Nigeria has the most significant relationship to the conduct at issue, which it identified as “Pfizer’s failure to inform the children or their parents about the potential problems with Trovan, and the administration of Trovan and low dosage of Ceftriaxone.” Adamu, 399 F.Supp.2d at 503 (citations omitted). It reasoned that “the Nigerian contacts to this litigation are stronger than Connecticut’s” and noted in particular that both the plaintiffs’ injuries and Pfizer’s alleged conduct occurred in Nigeria, that the plaintiffs were Nigerian residents, and that “the parties’ relationship is centered” in Nigeria. Id. It determined that most of the factors of Section 145(2) point toward applying Nigerian law and that the “sole basis” for the applicability of Connecticut law was that “Pfizer performed research and development with respect to Trovan and planned the experiment in Connecticut.” Id. For these reasons, it concluded that Nigeria’s interests were superior and that its law should apply. Id.

Although the district court correctly identified some of the pertinent factors, it ultimately erred in its application of the “most significant relationship” test because it did not factor into its Section 145(2) analysis the integral factors set out in Section 6(2). It did not, for example, discuss “the relevant policies of the forum” or “the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue.” Restatement (Second) § 6(2)(b)-(c). Nor did it analyze what “justified expectations” existed that could have prompted Pfizer reasonably to believe that its conduct in Connecticut would not expose it to Connecticut law, or how Pfizer would have been disadvantaged by litigating these claims in Connecticut. Id. § 6(2)(d). Finally, the district court did not evaluate its own ability to determine and apply Connecticut, as opposed to Nigerian, law. Id. § 6(2)(g). For these reasons, we vacate the dismissal of the state law claims and remand to the district court for further consideration.

CONCLUSION

For the foregoing reasons, we REVERSE the judgments of the district court and REMAND for further proceedings.

dissenting:

The majority has undertaken to define a “firmly established” norm of international law, heretofore unrecognized by any American court or treaty obligation, on the basis of materials inadequate for the task. In deviating from our settled case law, the majority identifies no norm of customary international law, it creates a new norm out of whole cloth. Because the majority’s analysis misconstrues — rather than vindicates — customary international law, I respectfully dissent.

*192Proceeding with “extraordinary care and restraint,” Flores v. S. Peru Copper Corp., 414 F.3d 233, 248 (2d Cir. 2003), this Court has upheld jurisdiction under the Alien Tort Statute, 28 U.S.C. § 1350 (“ATS”), in only a handful of cases alleging violations of the most firmly established international law norms, see Kadic v. Karadzic, 70 F.3d 232, 241-43 (2d Cir. 1995) (genocide and war crimes); Amerada Hess Shipping Corp. v. Argentine Republic, 830 F.2d 421, 426 (2d Cir. 1987), rev’d on other grounds, 488 U.S. 428, 109 S.Ct. 683, 102 L.Ed.2d 818 (1989) (free passage of neutral ship in international waters); Filartiga v. PenaIrala, 630 F.2d 876, 878 (2d Cir. 1980) (state-administered torture). In Sosa v. Alvarez-Machain, the Supreme Court identified three such “paradigmatic” norms, namely “violation of safe conducts, infringement of the rights of ambassadors, and piracy.” 542 U.S. 692, 724, 124 S.Ct. 2739, 159 L.Ed.2d 718 (2004). Rather than declare that list exhaustive for purposes of the ATS, the Court held that “any claim based on the present-day law of nations [must] rest on a norm of international character accepted by the civilized world and defined with a specificity comparable to the features of the 18th-century paradigms we have recognized.” Id. at 725, 124 S.Ct. 2739. Accordingly, we are charged with “vigilant doorkeeping” when reviewing ATS claims to ensure that they rest on “a narrow class of international norms” comparable to the paradigms identified by the Supreme Court. Id. at 729, 124 S.Ct. 2739.

The majority identifies three criteria that must be satisfied before a violation of international law can be actionable under the ATS: that the norm is (1) specific and definable, (2) universally adhered to out of a sense of legal obligation, and (3) a matter of mutual concern, namely a matter “involving States’ actions performed towards or with regard to the other.” Flores, 414 F.3d at 249 (internal quotation and alterations omitted). I agree with the methodology used by the majority to determine whether a norm falls within the jurisdictional grant of the ATS, but I do not agree with their conclusion that a norm against non-consensual medical experimentation on humans by private actors is (1) universal and obligatory or (2) a matter of mutual concern.

The majority relies on eight sources of customary international law to support its determination that a norm against non-consensual medical experimentation on humans by private actors is universal and obligatory. However, this evidence falls far short of the quantum necessary to establish the existence of such a norm: (1) the International Covenant on Civil and Political Rights has been described by the Supreme Court as a “well-known international agreement! ] that despite [its] moral authority, ha[s] little utility,” in defining international obligations, Sosa, 542 U.S. at 734, 124 S.Ct. 2739, and moreover, it does not apply to private actors, such as the Defendant in this action; (2) the Council of Europe’s Convention on Human Rights and Biomedicine — a regional convention— was not ratified by the most influential nations in the region, such as France, Germany, Italy, the Netherlands, Russia and the United Kingdom, and it was promulgated on April 4, 1997, one year after the conduct at issue in this litigation; (3) the UNESCO Universal Declaration of Bioethics and Human Rights of 2005 and (4) the European Parliament Clinical Trial Directive of 2001 both also post-date the relevant time period by several years; (5) the Declaration of Helsinki issued by the World Medical Association, a private entity, and (6) the International Ethical Guidelines for Research Involving Human Subjects promulgated by the Council for International Organizations for Medical *193Sciences, another private entity, “express[] the sensibihties and the asserted aspirations and demands of some countries or organizations” but are not “statements of universally-recognized legal obligations,” Flores, 414 F.3d at 262; (7) states’ domestic laws, which, unsupported by express international accords, are not “significant or relevant for purposes of customary international law,” id. at 249; and (8) the so-called Nuremberg Code, a statement of principles that accompanied a criminal verdict, possesses at best “subsidiary” value as a judicial decision, Statute of the International Court of Justice art. 38, June 26, 1945, 59 Stat. 1031, 33 U.N.T.S. 993 (“ICJ Statute”). Taken together, this evidence falls short of charting the existence of a universal and obligatory international norm actionable against non-government actors under the ATS.1

In support of its determination that non-consensual medical experimentation by private actors is a matter of mutual concern, the majority reasons that non-consensual medical experiments breed distrust of medical interventions and thereby accelerate the spread of infectious diseases across international borders. It is not enough, however, that tortious conduct could create some sort of international consequence. In order for conduct to be a matter of mutual concern, it must “threaten[ ] serious consequences in international affairs.” Sosa, 542 U.S. at 715, 124 S.Ct. 2739. Such is the case when an ambassador is assaulted, for example, because the assault “impinge[s] upon the sovereignty of the foreign nation and if not adequately redressed could rise to an issue of war.” Id. Non-consensual medical experimentation by private actors simply does not present the same grave risk of serious consequences in international affairs and is therefore not a matter of mutual concern.

For these reasons, I conclude that non-consensual medical experimentation by private actors, though deplorable, is not actionable under international law and would therefore affirm the district court’s dismissal of Plaintiffs’ complaints.

DISCUSSION

I. Universal and Legally Obligatory Adherence

In order for a principle to become a norm of customary international law, states must universally abide by it out of a sense of legal obligation, and not merely aspiration. See Flores, 414 F.3d at 248. It might seem obvious, but before one can determine whether a principle is universally followed, one must define the principle in question. Like domestic law, international law is not a monolith — a unitary set of rules applying indiscriminately to all actors that come within its reach. To the contrary, international law consists of rules that govern only states, rules that apply to private parties — individuals and corporations — and other rules that regulate both evenhandedly. See, e.g., Restatement (Third) of Foreign Relations of the United States § 101 (1987) (“Restatement (Third)”). As a result, the Supreme Court has required courts deciding whether a principle is a customary international law norm to consider “whether international law extends the scope of liability for a violation of a given norm to the perpetrator being sued, if the defendant is a private actor such as a corporation or individual.” Sosa, 542 U.S. at 732 n. 20, 124 S.Ct. 2739; see also id. at 760, 124 S.Ct. 2739 (Breyer, J., concurring) (“The norm must *194extend liability to the type of perpetrator (e.g., a private actor) the plaintiff seeks to sue”).

The majority lists the norm at issue here as the prohibition of “medical experimentation on non-consenting human subjects,” Maj. Op. at 174-75, and proceeds to analyze that norm without regard to the alleged violator, see id. at 174-88. Put another way, the majority’s analysis would be no different if Plaintiffs had sued the Nigerian government, instead of, or in addition to, Pfizer. Such a broad, simplified definition ignores the clear admonitions of the Supreme Court — and conflicts with prior decisions of this Court — that a customary international law norm cannot be divorced from the identity of its violator. The majority’s analysis omits this critical consideration. As a result, the majority opinion presents only half of the equation. To my mind, the majority should have asked whether customary international law prohibits private actors from medical experimentation on non-consenting human subjects. That question must be answered in the negative.

A. The Majority’s Sources of Customary International Law

In Flores, we explained some of the difficulties inherent in determining what offenses violate customary international law:

Customary international law is discerned from myriad decisions made in numerous and varied international and domestic arenas. Furthermore, the relevant evidence of customary international law is widely dispersed and generally unfamiliar to lawyers and judges. These difficulties are compounded by the fact that customary international law ... does not stem from any single, definitive, readily-identiflable source.

414 F.3d at 247-48. We have consistently looked to the ICJ Statute as the starting point for determining the proper sources of international law. See, e.g., id. at 250-51; United States v. Yousef, 327 F.3d 56, 100-03 (2d Cir. 2003). That statute lists: (1) “international conventions, whether general or particular, establishing rules expressly recognized by the contesting states”; (2) “international custom, as evidence of general practice accepted as law”; (3) “the general principles of law recognized by civilized nations”; and, in certain circumstances (4) “judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law.” ICJ Statute art. 38.

The ability to pick and choose from this seemingly limitless menu of sources presents a real threat of “creative interpretation.” Flores, 414 F.3d at 248; see also Amerada Hess, 830 F.2d at 429 (Kearse, J., dissenting). To mitigate this risk, and to prevent courts from becoming “roving commission[s],” Flores, 414 F.3d at 262, we have, in our cases, methodically assessed the weight and relative influence of not only each class of sources listed in the ICJ Statute, but many individual sources within each class. The near-infinite list of international law sources makes adherence to this precedent of paramount importance, for our analysis demonstrates that not every source of international law carries equal weight.

Instead of following and applying our framework, the majority substitutes in its place a compelling narrative. Over the course of only a few pages, the majority employs several sources that it believes demonstrate a customary norm against medical experimentation by non-state entities and weaves them together to reach its conclusion. See Maj. Op. at 175-85. Nowhere does the majority examine these *195sources in the context required by Sosa. The majority does not discuss the weight of these sources, how they collectively demonstrate a customary norm, or how evidence supporting that norm compares with our ATS precedent. Had they done so, I am hopeful that my colleagues would reach the same conclusion that I do — that medical experimentation by private actors, while reprehensible, is not actionable under international law.

1. Treaties & Conventions

In Flores, we noted that treaties are the strongest evidence of customary international law because they “create legal obligations akin to contractual obligations on the States parties to them.” 414 F.3d at 256. “[W]e look primarily to the formal lawmaking and official actions of States ... as evidence of the established practices of States.” Yousef, 327 F.3d at 103. But not all treaties are equal. Although “[a]ll treaties that have been ratified by at least two States provide some evidence of the custom and practice of nations ... a treaty will only constitute sufficient proof of a norm of customary international law if an overwhelming majority of States have ratified the treaty.” Flores, 414 F.3d at 256. Moreover, the “evidentiary weight to be afforded to a given treaty varies greatly depending on (i) how many, and which, States have ratified the treaty, and (ii) the degree to which those States actually implement and abide by the principles set forth in the treaty.” Id. at 256-57. For instance, treaties ratified by the United States are of greater evidentiary value if they are either self-executing or executed through acts of Congress. See, e.g., id. at 257; Khulumani v. Barclay Nat’l Bank Ltd., 504 F.3d 254, 284 (2d Cir. 2007) (Katzmann, J., concurring).

The majority relies primarily on two treaties.

a. International Covenant on Civil and Political Rights

The International Covenant on Civil and Political Rights, Dec. 9, 1966, S. Exec. Doc. E, 95-2, 999 U.N.T.S. 171, 6 I.L.M. 368 (ratified by the United States June 8, 1992) (“ICCPR”) “guarantees a broad spectrum of civil and political rights to individuals within signatory nations.” United States v. Duarte-Acero, 296 F.3d 1277, 1282 (11th Cir. 2002). One of those rights- — to be free of non-consensual medical or scientific experimentation — is stated in Article 7.

The ICCPR is not appropriate evidence of customary international law for at least two reasons. First, the Supreme Court in Sosa explicitly described the ICCPR as a “well-known international agreement! ] that, despite [its] moral authority, ha[s] little utility under the standard set out in this opinion,” because the “United States ratified [it] on the express understanding that it was not self-executing and so did not itself create obligations enforceable in the federal courts.” 542 U.S. at 734-35, 124 S.Ct. 2739 (emphasis added).

Second, whatever limited weight the ICCPR has with regard to state action, it does nothing to show that a norm prohibiting involuntary medical experimentation applies to non-state entities. In citing its seemingly universal language, the majority overlooks the ICCPR’s operative section, which requires that “[e]ach State Party ... undertake! ] to respect and to ensure to all individuals within its territory and subject to its jurisdiction the rights recognized in the present Covenant.” ICCPR art. 2(1). Thus, despite its broad text, the ICCPR by its own terms, only governs “the relationship between a State and the individuals within the State’s territory.” Duarte-Acero, 296 F.3d at 1283. Because the ICCPR only creates obligations flow*196ing from a state to persons within its territory, a non-state actor cannot be said to have violated it. Thus, the ICCPR was relevant in Filartiga (decided before the Supreme Court limited its utility), in the context of state-administered torture of one of its citizens in contravention of one of the rights guaranteed by states in the ICCPR. See 630 F.2d at 884. But whatever its evidentiary value had Plaintiffs sued the Nigerian government, the ICCPR clearly has none where the question is whether international law includes a norm actionable against a private corporation.

b. Convention on Human Rights and Biomedicine

The second treaty cited by the majority is the Convention on Human Rights and Biomedicine, Apr. 4, 1997, E.T.S. No. 164 (the “Convention”), promulgated by the Council of Europe. See Maj. Op. at 183. Articles 52 and 16 3 of the Convention require that the subject of scientific research give his or her informed consent, which may be withdrawn at any time.

The first problem with the majority’s reliance on the Convention is that it is a regional agreement not signed by the most influential states in the region. Membership in the Council of Europe is limited to European states. See Statute of the Council of Europe, art. 4, May 5, 1949, E.T.S. No. 1. It is difficult to see how the Convention demonstrates the universality of the medical experimentation principle when its signatories are limited to one continent. The majority also notes that the Convention has been signed by thirty-four states, see Maj. Op. at 183, but overlooks that it has only been ratified by twenty-two, and a treaty only evidences the customs and practices of states that have ratified it. Flores, 414 F.3d at 256. Lastly, and perhaps more importantly, the Convention is lacking even as evidence of a European norm, since it has not been ratified by the more influential European states, including France, Germany, Italy, the Netherlands, Russia and the United Kingdom, and a treaty’s evidentiary value increases along with the influence in international affairs of the states that have ratified it. See id. at 257; Convention on Human Rights and Biomedicine, Chart of Signatures and Ratifications as of December 23, 2008, http://conventions.coe.int/ Treaty/Commun/ChercheSig.asp?NT = 164 & CM=8 & DF= 12/23/2008 & CL=ENG (“Convention Ratifications Chart”).

A second, more fundamental problem with the majority’s reliance on the Conven*197tion is that it was promulgated after the conduct at issue here. I know of no authority for an international ex post facto definition of the law of nations by later signed treaties. Cf. Vietnam, Ass’n for Victims of Agent Orange v. Dow Chem. Co., 517 F.3d 104, 118 (2d Cir. 2008) (“The United States did not ratify the 1925 Geneva Protocol until 1975. Accordingly, the Protocol cannot be said to have constituted ‘a treaty of the United States,’ 28 U.S.C. § 1350, during the period relevant to this appeal.”). Plaintiffs allege that the Trovan testing occurred in March and April of 1996, but the Convention was not opened for signature until April 4, 1997, and did not bind any state until Slovakia’s ratification on January 15, 1998. See Flores, 414 F.3d at 256(“A State only becomes bound by — that is, becomes a party to — a treaty when it ratifies the treaty.”); Convention Ratifications Chart. The Convention is without import to this inquiry. Two other post-1996 sources cited by the majority, the 2005 UNESCO Universal Declaration on Bioethics and Human Rights and the 2001 European Parliament Clinical Trial Directive share equal evidentiary irrelevance for the same reason.

2. Multinational Declarations of Principle

Plaintiffs and the majority cite several multinational declarations, including the World Medical Association’s Declaration of Helsinki and the International Ethical Guidelines for Research Involving Human Subjects promulgated by the Council for International Organizations of Medical Sciences (“CIOMS Guidelines”), as additional evidence that the prohibition against non-consensual medical experimentation applies to non-state actors. In doing so, the majority somehow overlooks our decisions in Flores and Yousef.

In Flores, plaintiffs sought to demonstrate customary international law by reference to multinational declarations. In response, we noted that a declaration, “which may be made by a multinational body, or by one or more States, customarily is a ‘mere general statement of policy [that] is unlikely to give rise to ... obligation[s] in any strict sense.’ ” 414 F.3d at 262 (quoting 1 Oppenheim’s International Law 1189 (Sir Robert Jennings & Sir Arthur Watts, eds., 9th ed. 1996)) (alterations in original). “Such declarations are almost invariably political statements — expressing the sensibilities and the asserted aspirations and demands of some countries or organizations — rather than statements of universally-recognized legal obligations.” Id. As a result, we concluded that “such declarations are not proper evidence of customary international law.” Id,, (emphasis added).

In Flores, the declarations we rejected were put forth by international governmental bodies, the Organization of American States and the United Nations Conference on Environment and Development. Id. at 263. Here, the two declarations embraced by the majority were put forward by entirely private organizations— hardly evidence of the state of international law. The Declaration of Helsinki was adopted by the World Medical Association, a group comprised not of member states, but of physicians and private national medical associations. “The World Medical Association (WMA) is an international organization representing physicians .... [and] has always been an independent confederation of free professional associations.” See The World Medical Association, “About the WMA,” http://www.wma.net/e/ about/index.htm. The express terms of the Declaration of Helsinki make it abundantly clear that it is hortatory, and not obligatory: “The World Medical Association (WMA) has developed the Declaration of *198Helsinki as a statement of ethical principles .... ” See World Med. Ass’n, Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects art. A(l), June 1964. Similarly, CIOMS is “an international non-governmental, non-profit organization.” CIOMS, “What is CIOMS?”, http://www.cioms.ch/ jan2008_whatis_eioms.pdf.

Treating these well-meaning, aspirational, but private, declarations as sources of international law runs counter to our observation in Yousef that “no private person — or group of men and women such as comprise the body of international law scholars — creates the law.” 327 F.3d at 102. This is so for good reason. As we have seen in our ATS jurisprudence, international custom gives rise to legally enforceable obligations. To include the political statements of private organizations in the select and conscribed group of sources capable of creating international law would enfranchise non-democratic, unaccountable entities with governmental authority. As a result, these declarations are “not proper evidence of customary international law.” Flores, 414 F.3d at 262.

The majority focuses its lens on one line in Filartiga for the proposition that a “declaration may by custom become recognized as laying down rules binding upon the States.” Maj. Op. at 177 (quoting Filartiga, 630 F.2d at 883). In Filartiga, we were discussing a United Nations declaration, which though not binding, “creates an expectation of adherence” because it “speeif[ies] with great precision the obligations of member nations.” 630 F.2d at 883. The declarations relied on by the majority were not put forth by a governmental body such as the United Nations but by wholly private organizations, incapable of creating legally binding obligations.

3. State Practice

The majority also points to the great number of states that, in their respective domestic laws, require informed consent in medical research. That many countries have prohibited private actors from conducting medical experiments or treatments without informed consent is certainly commendable and worthy of praise, but not “significant or relevant for purposes of customary international law.”4 See Flores, 414 F.3d at 249. For it is only when states prohibit domestic action as a result of “express international accords” that a wrong becomes a violation of customary international law. See Filartiga, 630 F.2d at 888 (quoting IIT v. Vencap, Ltd., 519 F.2d 1001, 1015 (2d Cir. 1975) (Friendly, J.)). No such international accord exists here.

Moreover, “substantive uniformity” among states’ domestic laws is only a starting point for demonstrating international custom through individual state practice, which should also reflect a “procedural” consensus among states on how that behavior should be prosecuted — criminally and civilly. See Sosa, 542 U.S. at 761-62, 124 S.Ct. 2739 (Breyer, J, concurring). As Justice Breyer noted in his Sosa concurrence, the states of the world have reached both substantive and procedural agreement with respect to only a handful of certain international law norms made actionable against non-state entities. See id.; Part 1(B) infra. Non-consensual medical testing is not among them.

4. The Nuremberg Code

The majority centers its analysis around the Nuremberg Code, but, in the process, *199critically misstates its genesis and status in international law. See Maj. Op. at 177-79. Because the Code is a sui generis source of international law, its context is vital to understanding what it is — and what it is not.

The Nuremberg trials are unquestionably one of this country’s greatest and most enduring contributions to the field of international law. As early as 1943, the Allied powers contemplated bringing Nazi war criminals to justice after the conclusion of the Second World War. At the October 1943 Moscow Conference, the United States, United Kingdom and Soviet Union issued a joint “Statement on Atrocities,” warning that:

At the time of granting of any armistice to any government which may be set up in Germany, those German officers and men and members of the Nazi party who have been responsible for or have taken a consenting part in the above atrocities, massacres and executions will be sent back to the countries in which their abominable deeds were done in order that they may be judged and punished according to the laws of these liberated countries and of free governments which will be erected therein.

Moscow Declaration Statement of Atrocities, Oct. 30, 1943, 9 U.S. Dept of State Bull. 310 (signed by President Roosevelt, Prime Minister Churchill and Premier Stalin). The statement added that German criminals “whose offenses have no particular geographical localization ... will be punished by joint decision of the government of the Allies.” Id.

Following victory in Europe and the surrender of Germany, the Allies executed the London Charter on August 8, 1945, establishing an International Military Tribunal to try the “major war criminals,” London Charter, Agreement for the Prosecution and Punishment of the Major War Criminals of the European Axis, art. 3, Aug. 8, 1945, 59 Stat. 1544, 82 U.N.T. S. 279, and leaving the door open for other war criminals to be tried in any other “national or occupation court” that might be established, id. art. 6. Alongside the London Charter, the Allies promulgated the Charter of the International Military Tribunal and formed a four-member tribunal with one member appointed by each of the Allies, with jurisdiction over “the major war criminals” accused of committing three crimes: crimes against peace,5 war crimes,6 and crimes against humanity.7 Charter of the International Military Tribunal, arts. 2, 6, Aug. 8, 1945, 59 Stat. 1544, 82 U.N.T.S. 279. It was the International Military Tribunal that conducted the *200celebrated trial that resulted in the convictions of 19 of 22 defendants, including high-ranking Nazi officials Hermann Goering, Rudolf Hess, and Karl Doenitz. See generally Robert H. Jackson, Final Report to the President on the Nuremberg Trials (Oct. 7, 1946). But the Nuremberg Code was adopted by a different tribunal in a different trial.

Four months after the London Charter established the International Military Tribunal, the Allied Control Council, the joint allied entity that governed post-war Germany, enacted Control Council Law No. 10, which authorized each of the occupying Allies, within its own “Zone of Occupation,” to arrest and prosecute “persons within such Zone suspected of having committed a crime,”8 subject to a right of first refusal by the International Military Tribunal. Allied Control Council Law No. 10 art. Ill, §§ 1, 3 (Dec. 20, 1945), in 1 Trials of War Criminals Before the Nuernberg Military Tribunals Under Control Council Law No. 10, XVTII (William S. Hein & Co., Inc. 1997) (1949), available at http:// www.loc.gov/rr/frd/Military_law/pdf/NT_ war-criminals_Vol-I.pdf (“1 Trials of War Criminals ”).

The first of the American trials arising under Control Council Law No. 10 was the “Medical Case” against German doctors. On October 25, 1946, the American Office of Military Government for Germany enacted General Order 68, constituting Military Tribunal 1, comprised of three American military judges and one alternate judge. Id. at 5. That same day, Brigadier General Telford Taylor, Chief of Counsel for War Crimes, signed an indictment in United States v. Karl Brandt, et al. charging 23 defendants with war crimes, crimes against humanity, and conspiracy, and charging 10 of the defendants with membership in the “SS,” an organization declared criminal by the International Military Tribunal. Id. at 8-18. These charges were premised, primarily, on the defendants’ forced medical experiments, which constituted war crimes when performed on prisoners of war, and crimes against humanity when conducted on Nazi concentration camp prisoners.

At the conclusion of the Medical Case, 16 of the 23 defendants were convicted of one or more of the charges, and seven were ultimately sentenced to death. Along with their verdict, the military judges enumerated ten principles that came to be known as the Nuremberg Code, the first of which states that in medical experiments, the “voluntary consent of the human subject is absolutely essential.” 2 Trials of War Criminals Before the Nuernberg Military Tribunals Under Control Council Law No. 10, 181 (William S. Hein & Co., Inc. 1997) (1949), available at 7 http://www.loc.gov/rr/frd/ Military_law/pdf/NT_war-criminals_Vol-II. pdf (“2 Trials of War Criminals ”).

My colleagues contend that the Code flowed naturally from the principles of law espoused in the London Charter. They are quite right, of course, that Control Council Law No. 10 was modeled after the London Charter and the American and International military tribunals shared largely the same general international law and procedural frameworks. The London Charter identified and defined certain international law offenses — Crimes Against Humanity, Crimes Against Peace, and War Crimes — while each of the twelve trials before the American military tribunal con*201cerned a unique and horrific context for the commission of those crimes, ranging from medical experimentation on prisoners to the use of slave labor. For example, the definitions of Crimes Against Humanity and War Crimes under which the Nazi doctors were tried in the Medical Case were virtually identical to those of the London Charter. However, the majority overlooks the fact that the Nuremberg Code dealt not with these general principles of law, but instead with the very specific issue of permissible medical experimentation. The ethical principles espoused in the Code had no forebears in either the London Charter or the judgment of the International Military Tribunal. They were developed exclusively in the Medical Case.

I recite this history not to suggest that the Nuremberg Code is not an extraordinary or groundbreaking document, but rather to demonstrate the difficulty inherent in measuring its evidentiary weight, as it does not fit neatly into any of the categories this Court has identified for sources of international law. For one thing, the Code was developed by the United States military and announced by an American military court. See United States v. Stanley, 483 U.S. 669, 687, 107 S.Ct. 3054, 97 L.Ed.2d 550 (1987) (Brennan, J., dissenting). Certainly, the Code is not a treaty and did not immediately bind any state. Under the framework of the ICJ Statute— and, accordingly, this Court — because it was part of a criminal verdict, its closest analogue is a judicial decision, but judicial decisions are only “subsidiary,” rather than primary, sources of customary international law. See ICJ Statute art. 38; Maj. Op. at 173-74. I agree with my colleagues that the Code has had significant import — influence that continues to this day. The Code surely has evidentiary value in our inquiry, but there is nothing to indicate that the Code establishes a norm of international law prohibiting non-consensual medical experimentation or treatment by private actors, or compensates for the virtually non-existent evidentiary value of the other sources cited by the majority.

Conscious of our obligation to measure the weight of the sources of international law in the aggregate, what is the sum of the sources that serve as the cornerstone of the majority’s conclusion? The ICCPR, characterized by the Supreme Court as being of “little utility,” Sosa, 542 U.S. at 734, 124 S.Ct. 2739, which, in any event, does not apply to private actors; a pair of private organizations’ declarations that our Circuit precedent tells us “are not proper evidence of customary international law,” Flores, 414 F.3d at 262; one regional convention and two multi-national declarations that post-date the critical time period and are thus completely irrelevant; states’ domestic laws untethered to any international agreement that we are told is not “significant or relevant for purposes of customary international law,” id. at 249, 414 F.3d 233; and the Nuremberg Code, a document whose evidentiary value is unclear.

Simply put, the evidence here does not compare with the sources put forward in the few cases where we have held a principle to be a norm of customary international law. Exercising “extraordinary care and restraint,” see id. at 248, we have only upheld ATS jurisdiction in cases where the evidence of customary international law was entirely overwhelming.9 In Filartiga, *202we were persuaded by the fact that the “international consensus surrounding torture has found expression in numerous international treaties and accords.” 630 F.2d at 883 (emphasis added). There, the State Department — “the political branch with principal responsibility for conducting the international relations of the United States,” Flores, 414 F.3d at 262—had expressly announced that the prohibition against torture had ripened into a norm of customary international law.10 Filartiga, 630 F.2d at 884. In Kadic, we observed that genocide was included in section 404 of the Restatement and that the Convention on the Prevention and Punishment of the Crime of Genocide had been ratified by more than 120 nations, including the United States, 70 F.3d at 240-42, while international criminalization of war crimes was established by four Geneva Conventions, ratified by more than 180 nations, including the United States, id. at 242-43. In Amerada Hess, it was similarly obvious that Argentina’s Falkland War attack on an American ship violated one of the oldest customary international law norms. 830 F.2d at 423-24. We cited a variety of international accords establishing the right of a neutral ship to free passage. Id. at 424. After tracing the norm to Blackstone, we concluded that it was “beyond controversy that attacking a neutral ship in international waters ... violates international law.” Id.

In those cases, the evidence of international acceptance of each norm with respect to each defendant was “clear and unambiguous.” Flores, 414 F.3d at 252. In each case, the nations of the world gathered to ratify in universal numbers treaties that specifically prohibited genocide, war crimes, torture, and attacks on neutral ships — not in generalized human rights agreements but in accords with those discrete norms as their exclusive subjects.

My colleagues contend that I look only to the presence (or, in this case, the absence) of a globally ratified treaty as the exclusive source of an international law norm. Far from it — we have held that customary international law “does not stem from any single, definitive, readily-identifiable source.” Id. at 248. However, the great weight of ATS jurisdiction must rest upon a foundation sturdy enough to support it. Just as it would be error to stubbornly require one source of sufficient strength to bear that burden on its own, the majority is equally mistaken in its attempt to employ a series of extraordinarily weak sources to secure a purported norm of customary international law. Our case law makes clear that even when viewed collectively, these sources are incapable of carrying the weight placed upon them by my colleagues.

B. Restatement § ¿OJ

Nor does Plaintiffs’ purported norm resemble the select few norms for which international law extends liability to private actors. Although the law of nations *203in general does not “confine[ ] its reach to state action,” see Kadic, 70 F.3d at 239, courts must still consider whether the specific norm at issue does. In Kadic, we noted that the Restatement (Third) of Foreign Relations Law of the United States differentiates between “those violations that are actionable when committed by a state11 and a more limited category of violations” that apply with equal force to private actors. Id. at 240 (citing Restatement (Third) §§ 404, 702). Section 404 of the Restatement authorizes universal criminal jurisdiction over non-state entities “for certain offenses recognized by the community of nations as of universal concern, such as piracy, slave trade, attacks on or hijacking of aircraft, genocide, war crimes, and perhaps certain acts of terrorism, even where [no other basis of jurisdiction] is present.”12 Universal jurisdiction, not to be confused with universal acceptance of a norm for ATS purposes, “permits a State to prosecute an offender of any nationality for an offense committed outside of that State and without contacts to that State.” Yousef, 327 F.3d at 103.

The plaintiffs in Kadic alleged that Radovan Karadzic, the “president” of the self-proclaimed republic of Srpska violated several international law norms, notably bans on genocide, war crimes and torture. 70 F.3d at 236-37. Treating Karadzic as a non-state actor, we reviewed not only the Restatement, but a host of relevant international accords, leading us to conclude that by their own terms, the norms prohibiting genocide and war crimes applied to private individuals, while torture and summary execution “are proscribed by international law only when committed by state officials or under color of law.” Id. at 241-43. We added that the “ ‘color of law’ jurisprudence of 42 U.S.C. § 1983 is a relevant guide to whether a defendant has engaged in official action for purposes of jurisdiction under the [ATS].” Id. at 245.

Five years later, we again determined whether an international law norm applied only to state actors. See Bigio v. Coca-Cola Co., 239 F.3d 440 (2d Cir. 2000). Building on Kadic, we held that ATS jurisdiction over a non-governmental entity requires the violation of a norm “listed as an ‘act of universal concern’ in § 404 or ... sufficiently similar to [those] acts for us to treat them as though they were incorporated into § 404 by analogy,” or conduct committed under color of law. Id. at 448. In affirming the district court’s dismissal, we determined that the act at issue — discriminatory expropriation of property — is much more like the acts listed in section 702 than those in section 404, and that the complaint did not allege that Coca-Cola acted in concert with Egyptian state offi*204cials. Id. at 447-49. However, unlike in Kadic, we saw no need to look beyond the Restatement to any sources of international law in order to conclude that the norm did not apply to non-state entities. Compare id. at 448, with Kadic, 70 F.3d at 241-43. It is equally clear that section 404 of the Restatement does not reveal a norm of customary international law prohibiting non-consensual medical experimentation by private actors.

To reiterate, section 404 lists only five specific acts for which universal criminal jurisdiction over private actors exists: piracy, genocide, slave trade, war crimes, and attacks on aircrafts. See also Vietnam Ass’n for Victims of Agent Orange, 517 F.3d at 116 (describing these five as comprising “the list of principles that may be said to have ripened into universally accepted norms of international law” (internal quotation marks omitted)). If anything, this Court has been even more stringent, holding that in spite of the Restatement, federal courts could not try an alleged airline bomber under customary international law principles of universal jurisdiction.13 See Yousef, 327 F.3d at 103-08. Regardless, there is no dispute that none of the five acts in section 404 encompasses non-consensual medical experimentation. Instead, Plaintiffs argue that it is “sufficiently similar” to those acts to support its application to a private corporation.14 See Bigio, 239 F.3d at 448. This Court has never had occasion to consider what types of acts are “sufficiently similar” to the section 404 acts except to conclude in Bigio that discriminatory expropriation was not among them. Id. For similar reasons, neither is non-consensual medical experimentation.

Universal jurisdiction originated with prosecutions of piracy more than 500 years ago. See Yousef, 327 F.3d at 104; United States v. Lei Shi, 525 F.3d 709, 723 (9th Cir. 2008). As we explained in Yousef, piracy is universally punishable not because it is uniquely heinous but “because of the threat that piracy poses to orderly transport and commerce between nations and because the crime occurs statelessly on the high seas.” 327 F.3d at 104. By 1822, it was beyond “doubt ... that vessels and property in the possession of pirates may be lawfully seized on the high seas by [any] person, and brought in for adjudication.” United States v. the La Jeune Eugenie, 26 F. Cas. 832, 843 (C.C.D.Mass. 1822) (No. 15,551); see also United States v. Smith, 18 U.S. 153, 5 Wheat. 153, 163, 5 L.Ed. 57 (1820) (Story, J.) (discussing the bases for universal jurisdiction over piracy).

*205Private actors trading slaves (as opposed to those engaging in slavery in general) are subject to universal criminal jurisdiction because the early treaties that formed the basis for customary international law considered the slave trade akin to piracy. For example, the 1841 Treaty of London provided that:

Their Majesties the Emperor of Austria, the King of Hungary and Bohemia, the King of Prussia, and the Emperor of all the Russians, engage to prohibit all trade in slaves, either by their respective subjects, or under their respective flags, or by means of capital belonging to their respective subjects; and to declare such traffic piracy. Their Majesties further declare that any vessel which may attempt to carry on the Slave Trade, shall, by that fact alone, lose all right to the protection of their flag.

Treaty for the Suppression of the African Slave Trade art. I, Dec. 20, 1841, 92 Con-sol. T.S. 437 (emphasis added), reprinted in M. Cherif Bassiouni & Edward M. Wise, Aut Dedere Aut Judicare: The Duty to Extradite or Prosecute in International Law 132-33 (1995); see also Kenneth C. Randall, Universal Jurisdiction Under International Law, 66 Tex. L.Rev. 785, 798 (1988) (“Currently, states can recognize universal jurisdiction over slave trading by ... customary law.”). Although we declined to hold in Yousef that the principle had ripened into a customary norm, attacks on airliners logically fit into this class because, like the high seas, airspace is stateless and extraterritorial.

After World War II, universal criminal jurisdiction was extended to private actors — including many of the Nazi defendants prosecuted under Control Council Law No. 10 — accused of crimes against humanity such as war crimes and genocide because, like piracy, “ ‘there is ... a lack of any adequate judicial system operating on the spot where the crime takes place— in the case of piracy it is because the acts are on the high seas and in the case of war crimes because of a chaotic condition or irresponsible leadership in time of war.’ ” Yousef, 327 F.3d at 105 (quoting Willard B. Cowles, Universality of Jurisdiction Over War Crimes, 33 Cal. L.Rev. 177, 194 (1945)); see also Flores, 414 F.3d at 244 n. 18 (“Customary international law rules proscribing crimes against humanity, including genocide, and war crimes, have been enforceable against individuals since World War II.”).

In Yousef, we concluded that these acts share two common traits: they “(1) are universally condemned by the community of nations, and (2) by their nature occur either outside of a State or where there is no State capable of punishing, or competent to punish, the crime.” 327 F.3d at 105.

Non-consensual medical experimentation is not “sufficiently similar” to these crimes to warrant its incorporation into section 404 by analogy. Plaintiffs acknowledge that the acts listed in section 404 share “a particular quality of crossing international boundaries,” a quality that they argue that medical experimentation shares “because of the universal uses of medical research and the common practice of physicians to travel to crisis areas to deliver humanitarian aid.” But the mere crossing of an international border does not give rise to universal jurisdiction over non-state actors. We made this clear in Yousef where we rejected universal jurisdiction over an individual accused of bombing of an aircraft leaving the Philippines for Japan. 327 F.3d at 98, 103. As we held, universal criminal jurisdiction over private actors is only appropriate for acts which, “by their nature,” are beyond state sovereignty. Id. at 105. Here, Pfizer’s alleged actions occurred exclusively within Nigeria, and *206medical experimentation is not a crime which, by its nature, is incapable of state punishment. Plaintiffs’ argument to the contrary is belied by the state and federal civil and criminal actions pending against Pfizer in Nigeria. See Maj. Op. at 171-72.

As in Bigio, medical experimentation more closely resembles the acts for which only state actors may be held responsible. Plaintiffs compare medical experimentation with slavery. Yet, under the Restatement, while anyone may be prosecuted for engaging in the slave trade, slavery itself is only actionable against state actors. See Restatement (Third) § 702(b) (“A state violates international law if, as a matter of state policy, it practices, encourages, or condones ... slavery ... ”). Medical experimentation resembles slavery in its grievous exploitation of unconsenting and unwilling subjects; it also resembles torture in its infliction of horrific physical and emotional pain. However, both the Restatement and this Court have recognized that the norm against torture reaches only state actors. See Kadic, 70 F.3d at 243-44; Restatement (Third) § 702(d); see also Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment art 1, Dec. 10, 1984, S. Treaty Doc. No. 100-20, 1465 U.N.T.S. 85 (“CAT”) (defining torture as being “inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity.”).15

One of the fundamental attributes of sovereignty is a state’s authority to exercise criminal jurisdiction over persons accused of committing crimes within its territory. The crimes listed in section 404 are not the extraordinary exceptions because they are singularly reprehensible or deserving of condemnation. Few would argue that piracy, for which private actors may be prosecuted universally but which requires neither an act of violence nor the infliction of physical injury,16 is more heinous than torture or slavery, practices made actionable only against state entities. Rather, by definition, these crimes occur in locations where, or during times when, sovereignty, and a fortiori criminal jurisdiction, are incapable of being exercised. Because medical experimentation is entirely intra national and fully subject to domestic criminal jurisdiction, it is not “sufficiently similar” to those acts listed in section 404, and cannot be incorporated by analogy as to reach private, non-state actors.

The defendants in the Medical Case were not charged with conducting non-consensual medical tests per se. Rather, those tests, when conducted on prisoners of war and members of a discrete civilian population imprisoned in concentration camps, constituted “war crimes” and “crimes against humanity,” offenses for which customary international law has imposed individual responsibility. See Flores, 414 F.3d at 244 n. 18. Unlike the Defendant in this action, the Nazi doctors convicted by the American military tribunal were not private actors. Each convicted defendant held a position of authority in either the medical services or the military of the Third Reich.17 See 1 Trials of War *207Criminals 29. Moreover, the atrocities for which they were convicted victimized state prisoners in state-administered concentration camps, according to the Indictment, “for the benefit of the German Armed Forces.” Id. at 11-14. It is difficult to imagine a more egregious example of the violation of a customary international law norm or a more appropriate case for ATS jurisdiction.

The majority today authorizes the exercise of ATS jurisdiction over an entirely private corporation for violating a previously unrecognized norm of international law. In doing so, my colleagues accept proof far weaker than in any other case where this Court has identified a norm of customary international law, and, apparently, overlook the fact that this purported norm in no way resembles those few norms enforceable against private entities. When tasked by the Supreme Court with “vigilant doorkeeping” to ensure that the list of actionable international norms remains “narrow,” Sosa, 542 U.S. at 729, 124 S.Ct. 2739, we must be no less demanding than we have been in the past. Under that standard, the evidence put forward by Plaintiffs does not establish a norm of customary international law actionable against private actors. I believe that the majority’s decision departs from our settled case law and lowers considerably our previously high bar for ATS jurisdiction.

II. Mutuality

There are many principles on which most states of the world community agree. Most find support and enforcement in the richly diverse legal systems in place around the globe. But universal acceptance as a normative principle is not enough to gain entrance into the “law of nations.” The norm must not only be universal, it must touch on matters that are “of mutual, and not merely several, concern.” Filartiga, 630 F.2d at 888. Matters are of mutual concern when they “affectf] the relationship between states or between an individual and a foreign state, and [are] used by those states for their common good and/or dealings inter se.” IIT, 519 F.2d at 1015. On the other hand, matters of several concern are those “in which States are separately and independently interested.” Flores, 414 F.3d at 249. For example, as we noted in Flores, “murder of one private party by another, universally proscribed by the domestic law of all countries ... is not actionable under the [ATS] as a violation of customary international law because ‘the nations of the world’ have not demonstrated that this wrong is of mutual, and not merely several, concern.” Id. (quotation marks omitted). The majority concludes that non-consensual medical experimentation by one private party on another is a matter of mutual concern. I disagree.

We have consistently held that the best evidence that states consider a matter to be of mutual concern is the fact that they have agreed to be bound “by means of express international accords.” Filartiga, 630 F.2d at 888; see Flores, 414 F.3d at 249; Khulumani, 504 F.3d at 274 n. 7 (Katzmann, J., concurring). The majority points to the ICCPR, the Convention on Human Rights and Biomedicine, and the 2001 Clinical Trial Directive as evidence that “States throughout the world have entered into ... express and binding international agreements prohibiting nonconsensual medical experimentation.” See *208Maj. Op. at 185. But those agreements fail to demonstrate mutuality for the same reason they fail to demonstrate universality — the ICCPR does not address acts by non-state actors and the other two were not in force at the time of the alleged misconduct. Whatever international consensus has been reached as to non-consensual medical experimentation by private actors has not yet “found expression in numerous treaties and accords,” of. Filartiga, 630 F.2d at 883. The majority cites no worldwide, multi-continental, universally applicable “Convention Against Medical Experimentation,” because, at the moment, none exists. That fact alone distinguishes this case from Filartiga, Amerada Hess, and Kadic.

In the absence of a binding global treaty, the majority seeks to demonstrate mutuality of concern by describing the downstream effects of non-consensual medical experimentation. In essence, the majority contends that non-consensual medical experiments feed distrust among their victims, which, in turn, engenders a general reluctance to seek future medical attention or vaccination, which, in turn, helps accelerate the spread of infectious diseases across international borders. See Maj. Op. at 186-87. Indeed, I would concede that the majority may be quite right. But a smaller, more interdependent world community has not been employed by the Supreme Court (or any other court to my knowledge) to convert claims such as those presented here into violations of the law of nations. In fact, the majority’s theory would be no different when evaluating the medical malpractice of Pfizer’s research physicians or the strict products liability for its allegedly defective drug, but malpractice and products liability are among the quintessential subjects of domestic law.

It is not enough that a wrong could create international ramifications; in order for it to be a matter of mutual concern, it must “threaten!] serious consequences in international affairs.” Sosa, 542 U.S. at 715, 124 S.Ct. 2739. The Supreme Court listed three historical mutual wrongs as guideposts to frame this inquiry: infringement of the rights of ambassadors, the violation of safe conducts and piracy. Id. at 715, 720, 124 S.Ct. 2739. An assault against an ambassador “impinged upon the sovereignty of the foreign nation and if not adequately redressed could rise to an issue of war.” Id. at 715, 124 S.Ct. 2739. The 18th century safe-conduct document was the historical equivalent of the modern passport, “which entitles a bearer with a valid visa to safe passage to, within, and out of a foreign land pursuant to a treaty or an agreement negotiated by his or her sovereign and the host sovereign.” Taveras v. Taveraz, All F.3d 767, 773 (6th Cir. 2007) (quoting Thomas H. Lee, The Safe-Conduct Theory of the Alien Tort Statute, 106 Colum. L.Rev. 830, 874 (2006)). Thus, “the purpose of the doctrine of safe conducts under the law of nations is to protect the safety and security of the person and property of the journeying alien bearing the safe conduct privilege (and consequently to preserve commercial and diplomatic relationships between the alien’s host and home countries).” Id. at 773-74. This is still true today — a passport issued by the United States contains an official request from the Secretary of State to an authority of another sovereign state: “The Secretary of State of the United States of America hereby requests all whom it may concern to permit the citizen/national of the United States named herein to pass without delay or hindrance and in case of need to give all lawful aid and protection.” Breaches of customary international law impair the normal expectations that nations have in dealing with other nations. They must threaten serious consequences in international affairs because the norms were, and *209still are, the foundation for states’ formal relationships with one another.

Piracy does not fit squarely with the other two Sosa historical paradigms, but the threat to international affairs posed by piracy needs no detailed exegesis. Suffice it to say that one of the young Republic’s first military tests was its campaign against the Barbary Pirates, see, e.g., Act For the Protection of the Commerce and Seamen of the United States Against the Tripolitan Cruisers, ch. IV, § 2, 2 Stat. 129, 130 (1802) (authorizing President Jefferson to instruct the armed forces to “seize and make prize of all vessels, goods and effects, belonging to the Bey of Tripoli ... and also to cause to be done all such other acts of precaution or hostility as the state of war will justify, and may, in his opinion, require.”), and piracy continues to threaten serious consequences in international affairs today, see S.C. Res. 1851, ¶ 2, U.N. Doc. S/RES/1851 (Dec. 16, 2008) (calling upon states “to take part actively in the fight against piracy and armed robbery at sea off the coast of Somalia”).

We have accepted no lesser showing in our case law. The threat posed by genocide is so great that states are empowered to request “the competent organs of the United Nations to take such action under the Charter of the United Nations as they consider appropriate for the prevention and suppression of acts of genocide.” Convention on the Prevention and Punishment of the Crime of Genocide art. 8, Dec. 9, 1948, 102 Stat. 3045, 78 U.N.T.S. 277. The Geneva Conventions collectively establish, and obligate contracting parties to follow, the laws of war — almost by definition a matter of international affairs. See Kadic, 70 F.3d at 242-43. On the other hand, because international law does not define torture to include acts by private entities, torturous conduct by non-state actors — while criminalized domestically — is not a matter of mutual concern. Id. at 243-44.

Demonstrating that a wrong is a matter of mutual concern must necessarily be difficult. The Supreme Court has only opened the door for ATS jurisdiction over a “narrow set of violations of the law of nations, admitting of a judicial remedy and at the same time threatening serious consequences in international affairs.” Sosa, 542 U.S. at 715, 124 S.Ct. 2739. The nations of the world have not yet demonstrated that non-consensual medical experimentation by non-state actors “is of mutual, and not merely several, concern, by means of express international accords.” Filartiga, 630 F.2d at 888. Nor does it threaten serious consequences in international affairs in the same manner or to the same extent as the historical paradigms listed by the Supreme Court or their modern counterparts identified by this Court. Without either showing, I cannot agree with the majority that non-consensual medical experimentation by private actors is a matter of mutual concern.

III. State Action

The fact that medical experimentation by private actors is not a subject of customary international law does not end the inquiry. If international law supports state liability but not private liability, a private actor may still be liable if he or she “acted under color of law.” In that regard, we are told to employ our 42 U.S.C. § 1983 jurisprudence in the inquiry. See Bigio, 239 F.3d at 448; Kadic, 70 F.3d at 245. As an initial matter, this requires that the law of nations includes a norm actionable against states, which, in the instant case, is far from certain. But even assuming, for argument’s sake, that international law prohibits states from conducting non-consensual medical tests, Plaintiffs *210have not demonstrated that Pfizer acted under the color of law.

This issue requires a bit of procedural context. In 2002, Pfizer moved to dismiss Plaintiffs’ complaint in Abdullahi on the grounds that (1) Plaintiffs had not alleged that Pfizer was a state actor, and (2) the alternate ground of forum non conveniens. See Abdullahi v. Pfizer, Inc., No. 01 Civ. 8118, 2002 WL 31082956, at *12 (S.D.N.Y. Sept. 17, 2002). Judge Pauley granted the forum non conveniens motion, but denied the state action motion, concluding that Plaintiffs “sufficiently allege[d] that the former Nigerian government and Pfizer were joint participants in the Trovan treatment.” Id. at *6. Plaintiffs appealed the district court’s dismissal, and Pfizer cross-appealed from the court’s denial of its motion to dismiss on state action. See Abdullahi v. Pfizer, Inc., 77 Fed.Appx. 48 (2d Cir. 2003). On appeal, we vacated the district court’s judgment of dismissal, and did not reach Pfizer’s cross-appeal, noting that our intervening decision in Flores might have some application on remand. Id. at 53. Back before Judge Pauley, Pfizer filed a new motion to dismiss, arguing that Plaintiffs failed to state a claim under the substantially different ATS landscape which now included the Supreme Court’s decision in Sosa and our decision in Flores. See Abdullahi v. Pfizer, Inc., No. 01 Civ. 8118, 2005 WL 1870811, at *3 (S.D.N.Y. Aug. 9, 2005). Both of these decisions made clear that the identity of the defendant is a critical component of whether a principle is a norm of customary international law. Without addressing or affirming its previous conclusion finding sufficient allegations of state action, the district court granted Pfizer’s motion to dismiss, holding that medical experimentation was not actionable under the law of nations. Id. at *18. On appeal to this Court, both parties addressed the issue of state action in their briefs. The majority concludes that Plaintiffs’ allegations of state action were sufficient to defeat a motion to dismiss. See Maj. Op. at 187-88. I cannot agree.

In their twin complaints, which total 628 paragraphs, Plaintiffs make only four allegations concerning the role of the Nigerian government in the Trovan experiments: (1) in order for the FDA to authorize the export of Trovan, “Pfizer obtained the required letter of request from the Nigerian government”; (2) the government “arrang[ed] for Pfizer’s accommodation in Kano”; (3) the government acted “to silence Nigerian physicians critical of [Pfizer’s] test”; and (4) the government “assigned] Nigerian physicians to assist in the project.”18 Elsewhere in their complaints, Plaintiffs note in eonclusory fashion that a Nigerian doctor did not publicly object to the Trovan study because it “seemed to have the backing of the Nigerian government.”

In their brief to this Court, Plaintiffs seek to bolster their complaints by describing the role of “Nigerian government doctors” at the allegedly government-owned hospital that hosted the study. However, the portions of the complaints that they cite do not support their contentions. Nowhere in their complaints did Plaintiffs allege that the hospital was, in fact, government owned or administered, nor did they allege that the four Nigerian doctors working with Pfizer were employed by the government, and our review of a decision to grant a motion to dismiss “is limited to the facts as asserted within the four corners of the complaint” and any *211attached documents. McCarthy v. Dun & Bradstreet Corp., 482 F.3d 184, 191 (2d Cir. 2007).

These bare allegations are plainly insufficient to survive a motion to dismiss for lack of state action. The Supreme Court’s case law on state action is hardly a model of clarity, but certain principles are well-settled. As a threshold matter, the conduct alleged attributable to the state must be defined with the requisite specificity. “When analyzing allegations of state action, we begin ‘by identifying the specific conduct of which the plaintiff complains,’ ” Tancredi v. Metro. Life Ins. Co., 316 F.3d 308, 312 (2d Cir. 2003) (quoting Am. Mfrs. Mut. Ins. Co. v. Sullivan, 526 U.S. 40, 51, 119 S.Ct. 977, 143 L.Ed.2d 130 (1999)), and in most cases, a finding of state action “must be premised upon the fact that the State is responsible ” for that specific conduct, Horvath v. Westport Library Ass’n, 362 F.3d 147, 154 (2d Cir. 2004) (internal quotation omitted). Determining state action in these cases “requires tracing the activity to its source to see if that source fairly can be said to be the state.” Leshko v. Servis, 423 F.3d 337, 340 (3d Cir. 2005); see also Hadges v. Yonkers Racing Corp., 918 F.2d 1079, 1082-83 (2d Cir. 1990). As we recently stated, when confronted with a motion to dismiss, it “is not enough ... for a plaintiff to plead state involvement in some activity of the institution alleged to have inflicted injury upon a plaintiff; rather, the plaintiff must allege that the state was involved with the activity that caused the injury giving rise to the action.” Sybalski v. Indep. Group Home Living Program, Inc., 546 F.3d 255, 257-58 (2d Cir. 2008) (internal quotations omitted).

Here, that activity was not, as the majority apparently concludes, conducting the Trovan trials in general, but rather administering the drug without informed consent. Although Plaintiffs allege that the Nigerian government requested the import of Trovan and arranged for Pfizer’s accommodations and some medical staff in Kano, they do not allege that the government or any government employee played any role in either administering Trovan without consent or deciding to do so in the first instance. The Supreme Court has described “the typical case raising a state-action issue” as one in which “a private party has taken the decisive step that caused the harm to the plaintiff, and the question is whether the State was sufficiently involved to treat that decisive conduct as state action.” NCAA v. Tarkanian, 488 U.S. 179, 192, 109 S.Ct. 454, 102 L.Ed.2d 469 (1988). Plaintiffs have not alleged any facts that would indicate that the answer here is “yes.”

Plaintiffs’ complaints are more noteworthy for what they do not allege than what they do; They have not suggested that Pfizer was exercising any delegated state authority, cf. West v. Atkins, 487 U.S. 42, 108 S.Ct. 2250, 101 L.Ed.2d 40 (1988), or that the Nigerian government “knowingly accepted] the benefits derived from [the unlawful] behavior,” Tarkanian, 488 U.S. at 192, 109 S.Ct. 454. Plaintiffs have not alleged that Pfizer conspired with government officials to deprive the subjects of their rights, cf. Fries v. Barnes, 618 F.2d 988, 991 (2d Cir. 1980), nor have they alleged that the Nigerian government exercised any coercive power over Pfizer, cf. Brentwood Acad. v. Tenn. Secondary Sch. Athletic Ass’n, 531 U.S. 288, 296, 121 S.Ct. 924, 148 L.Ed.2d 807 (2001). In fact, Plaintiffs did not allege that any Nigerian government officials even knew about the non-consensual tests, because if Nigerian government doctors were somehow involved in the study, Plaintiffs did not specify what role, if any, they played.

The case of Beanal v. Freeport-McMoRan, Inc., 969 F.Supp. 362 (E.D.La. 1997), *212aff'd 197 F.3d 161 (5th Cir. 1999), is instructive. In Beanal, plaintiffs seeking to recover under the ATS sought to establish state action on the basis of the Indonesian military’s involvement in allegedly actionable conduct. The court rejected that argument, holding that plaintiffs had not “alleged whether the military personnel helped enforce Freeport’s policies or merely observed ... the violative conduct.” Id. at 378. Broad conclusory statements of state involvement are not sufficient to establish state action; “there must be some allegation indicating that the troops jointly cooperated in the conduct, jointly participated in the conduct, influenced the conduct or played an integral part in the deprivation of human rights.” Id. at 379. The same is true here.19 Plaintiffs’ allegations are inadequate.

Even without alleging that the State “coerced or even encouraged” the act complained of, Plaintiffs can still survive a motion to dismiss if “the relevant facts show pervasive entwinement to the point of largely overlapping identity between the State and the entity that the plaintiff contends is a state actor.” Horvath, 362 F.3d at 154 (quotation omitted). This line of cases revolves around the relationship between the state and the actor, as opposed to the specific act. Showing “overlapping identity” is highly uncommon, and most often arises where a private actor is performing one of the few functions traditionally and exclusively reserved to the state or is controlled by a state entity. State assistance by itself is insufficient — the relevant question is whether the decision-makers were ostensibly state actors. We answered that question in the affirmative in Horvath, where half of the putatively private defendant’s trustees were state appointees. Id. at 153. But the assistance alleged by Plaintiffs — helping to procure a ward in a hospital and arranging for the assistance of a handful of doctors — is not enough to clear this hurdle. Using government property, government staff, and even government funds does not make a private entity a state actor when its decisions are made independently of the state. See Yeo v. Town of Lexington, 131 F.3d 241, 254 (1st Cir. 1997) (en banc).

Plaintiffs’ generalized allegations (unsupported by factual allegations) that the government acted to silence critics of the test are no more helpful. They do not allege who these government officials were, how they acted to silence critics, or when in the sequence of events this conduct occurred. Such a “merely conclusory allegation that a private entity acted in concert with a state actor does not suffice to state a § 1983 claim against the private entity.” Ciambriello v. County of Nassau, 292 F.3d 307, 324 (2d Cir. 2002).

At most, Plaintiffs’ complaints alleged that the Nigerian government acquiesced to or approved the Trovan program in general without knowing its disturbing details. That it approved the program is hardly surprising — in the midst of a widespread epidemic, the Nigerian government likely welcomed help from every entity offering it, but “[mjere approval of or acquiescence in the initiatives of a private party is not sufficient to justify holding the State responsible for those initiatives.” *213Blum v. Yaretsky, 457 U.S. 991, 1004-05, 102 S.Ct. 2777, 73 L.Ed.2d 534 (1982). Plaintiffs have not demonstrated that Pfizer acted “under the color of law” such that it can be held liable for the Nigerian government’s alleged violation of the “law of nations.”

Plaintiffs’ allegations paint a vivid picture of the unspeakable pain and suffering of dozens of innocent children. The issue on this appeal, however, is not whether Pfizer’s alleged conduct was “wrong,” or even whether it is legally actionable, but whether it falls within both the “narrow class” of international norms for which ATS jurisdiction exists, and the even smaller subset of those norms actionable against non-state actors. Our Court and the Supreme Court have made it pellucidly clear that ATS jurisdiction must be reserved only for acts that the nations of the world collectively determine interfere with their formal relations with one another— including those rare acts by private individuals that are so serious as to threaten the very fabric of peaceful international affairs. I cannot agree with my colleagues that Pfizer’s alleged conduct poses the same threat or is so universally and internationally proscribed as to fit within that narrow class.

I respectfully dissent.