Supreme Court of Connecticut, 2026

Bryan v. Commissioner of Correction

Supreme Court of Connecticut · Decided September 1, 2026 · Mullins; McDonald; D’Auria; Ecker; Dannehy; Bright

Procedural history

Amended petition for a writ of habeas corpus, brought to the Superior Court in the judicial district of Tolland, geographical area number nineteen, and tried to the court, Bhatt, J.; judgment denying the petition, from which the petitioner, on the granting of certification, appealed. Reversed; judgment directed. Kayla R. Stephen, deputy assistant public defender, for the appellant (petitioner). Meryl R. Gersz, assistant state's attorney, with whom, on the brief, were Sharmese L. Walcott, state's attorney, and Christopher Alexy, senior assistant state's attorney, for the appellee (respondent).

Concurring in Part

************************************************ The “officially released” date that appears near the beginning of an opinion is the date the opinion will be published in the Connecticut Law Journal or the date it is released as a slip opinion. The operative date for the beginning of all time periods for the filing of postopinion motions and petitions for certification is the “officially released” date appearing in the opinion.

All opinions are subject to modification and technical correction prior to official publication in the Connecti- cut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports. In the event of discrepancies between the advance release version of an opinion and the version appearing in the Connecticut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports, the latest version is to be considered authoritative.

The syllabus and procedural history accompanying an opinion that appear in the Connecticut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports are copyrighted by the Secretary of the State, State of Connecticut, and may not be reproduced or distributed without the express written permission of the Commission on Official Legal Publications, Judicial Branch, State of Connecticut. ************************************************ DANNEHY, J., with whom BRIGHT, J., joins as to part II only, concurring in the judgment and dissenting in part. The majority concludes that the state violated Napue v. Illinois, 360 U.S. 264, 269–70, 79 S. Ct. 1173, 3 L. Ed. 2d 1217 (1959), by failing to correct purport- edly false or substantially misleading testimony given by Kingsley Minto and Attorney Richard Cohen at the crimi- nal trial of the petitioner, Carlton Bryan. In reaching that conclusion, the majority determines that the habeas court clearly erred in finding that the testimony of Minto and Cohen was neither false nor substantially misleading and, further, that the habeas court clearly erred in finding that, even if portions of Minto’s testimony were false or substantially misleading, any resulting misimpres- sion was corrected by Cohen’s testimony explaining the nature and extent of Minto’s understanding concerning his arrangement with the state. In my view, the majority reaches these conclusions only by failing to afford the deference required by our clearly erroneous standard of review. Because the habeas court’s factual findings are supported by evidence in the record and, therefore, are not clearly erroneous, I cannot agree that the petitioner established any Napue violation.

I conclude, however, that the petitioner’s conviction must nonetheless be reversed because he has established that the state failed to satisfy its obligations under Brady v. Maryland, 373 U.S. 83, 87, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963), and Giglio v. United States, 405 U.S. 150, 154, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972). Spe- cifically, the state was required, but failed, to disclose two interviews of Minto conducted by the Manchester Police Department, as well as certain benefits conferred on Minto, namely, the pretrial decision of the prosecutor, Senior Assistant State’s Attorney Donna Mambrino, not to sign arrest warrants supported by probable cause for Minto’s alleged involvement in armed robberies in West Hartford and Windsor. Because the state withheld impeachment evidence and that evidence was material, it violated its obligations under Brady and Giglio, and reversal is required. I therefore respectfully concur in the judgment and dissent in part.

In Napue, the United States Supreme Court held that a conviction knowingly “obtained through use of false evi- dence” violates the fourteenth amendment’s due process clause. Napue v. Illinois, supra, 360 U.S. 269. Testimony that is substantially misleading, even if “technically cor- rect,” is considered false under Napue. United States v. Bagley, 473 U.S. 667, 684, 105 S. Ct. 3375, 87 L. Ed. 2d 481 (1985); see id., 683–84; State v. Ouellette, 295 Conn. 173, 186, 989 A.2d 1048 (2010). “To establish a Napue violation, a defendant must show that the prosecution knowingly solicited false testimony or knowingly allowed it ‘to go uncorrected when it appear[ed].’ ” Glossip v. Oklahoma, 604 U.S. 226, 246, 145 S. Ct. 612, 221 L.

Ed. 2d 90 (2025), quoting Napue v. Illinois, supra, 269. “[The] essence of [a] Napue/Giglio violation is [the] lack of disclosure of [the] truth to [the] jury . . . .” Gomez v. Commissioner of Correction, 336 Conn. 168, 182, 243 A.3d 1163 (2020). “Whether a prosecutor knowingly presented false or misleading testimony . . . presents a mixed question of law and fact, with the habeas court’s factual findings subject to review for clear error and the legal conclusions that the court drew from those facts subject to de novo review.” (Internal quotation marks omitted.) Id., 175. Thus, whether testimony was false or substantially misleading and whether Mambrino cor- rected any misimpression are factual determinations entrusted in the first instance to the habeas court. See id. Our role is not to determine whether we would have interpreted the testimony differently but only whether the habeas court’s findings are unsupported by the evi- dence or otherwise leave this court with “the definite and firm conviction that a mistake has been committed.” (Internal quotation marks omitted.) Doyle v. Kulesza, 197 Conn. 101, 105, 495 A.2d 1074 (1985).

I conclude that the habeas court reasonably found that Minto’s testimony, considered in context and as a whole, was not false or substantially misleading concerning his arrangement with the state. The majority reaches the opposite conclusion by focusing almost exclusively on Minto’s repeated denials that he had received a “deal,” while discounting the testimony that immediately accom- panied those denials. On direct examination, after Minto denied that he had been promised “any deals,” Mambrino immediately asked whether he had been told that, if he testified, the state would advise the sentencing judge of his cooperation. Although Minto again responded that “[n]o deal was made,” Mambrino clarified the ques- tion, asking, “[n]o specific deal was made, but was that told to you?” Minto answered, “[y]es.” Thus, the jury was expressly informed that Minto had been told that Mambrino would make his cooperation known to the sentencing judge. Read fairly and in context, Minto’s repeated statements that there was “[n]o deal” reasonably referred to the absence of a negotiated plea agreement, sentencing recommendation, or promised disposition of his pending criminal charges—not to the absence of communications between the parties regarding his coop- eration. See Greene v. Commissioner of Correction, 330 Conn. 1, 17, 190 A.3d 851 (2018) (witness’ statement “cannot be divorced from the context surrounding it”), cert. denied sub nom. Greene v. Semple, 586 U.S. 1167, 139 S. Ct. 1219, 203 L. Ed. 2d 238 (2019).1 The habeas court reasonably interpreted the testimony in precisely that manner.

The remainder of Minto’s testimony at the petitioner’s criminal trial similarly does not compel the majority’s conclusion. Defense counsel thoroughly cross-examined Minto concerning his expectations arising from his tes- timony. Although Minto denied that anything had been promised to him, the jury also heard his acknowledgment at the conclusion of his testimony that he did have “con- siderations,” the details of which would be addressed later. Whether those answers were perfectly precise is not the question before us. The relevant question is whether the habeas court’s findings are unsupported by the evi- dence. On this record, I cannot conclude that they are. The habeas court also reasonably determined that, even if portions of Minto’s testimony could be viewed as incom- plete or potentially misleading, Mambrino corrected any possibly misleading testimony by calling Cohen, the attorney who represented Minto in connection with the Manchester robbery, immediately after Minto testified. See State v. Johnson, 345 Conn. 174, 208, 283 A.3d 477 (2022) (witness’ misleading testimony was cured by call- ing of another witness). Cohen explained what was meant by Minto’s statement that there was “[n]o deal” when Cohen testified that there was no written plea agreement, no agreement for any specific sentence, and no promise of any particular sentencing recommendation. He further testified, however, that Mambrino had informed him that, if Minto testified truthfully, she would advise the sentencing court of Minto’s cooperation and that Cohen hoped this would result in a more favorable disposition. Thus, the jury learned precisely the nature of Minto’s informal arrangement with the state.

The majority faults Cohen for describing this arrange- ment as an “ ‘expectation’ ” or “ ‘hope’ ” rather than as an “agreement . . . .” (Emphasis omitted.) Part II of the majority opinion. In doing so, however, it substitutes its own characterization of the evidence for that of the habeas court. The constitutional question is not whether every witness used identical terminology to describe the parties’ understanding. Rather, it is whether the jury was apprised of the facts bearing on the witness’ potential bias. See Napue v. Illinois, supra, 360 U.S. 269; Gomez v. Commissioner of Correction, supra, 336 Conn. 189–90.

The habeas court reasonably concluded that Cohen’s testimony accomplished exactly that by informing the jury that Minto expected favorable consideration because Mambrino had agreed to advise the sentencing judge of Minto’s cooperation. Indeed, Mambrino’s decision to call Cohen immediately after Minto underscores the absence of any effort to mislead the jury. Rather than allowing defense counsel to speculate about the nature of the par- ties’ understanding, Mambrino elicited testimony from Cohen, the attorney who had reached the understanding, and placed its essential terms before the jury. The habeas court therefore reasonably concluded that any potential ambiguity in Minto’s testimony had been clarified.

Because the record reasonably supports the habeas court’s determination that Minto’s testimony was not false or substantially misleading and, alternatively, that any potentially misleading testimony was corrected through Cohen’s testimony, I cannot conclude that those factual findings are clearly erroneous. The majority reaches a different result only by reweighing the evi- dence and drawing its own factual inferences from the trial record. Our standard of review does not permit that approach. See, e.g., State v. Michael J., 274 Conn. 321, 347, 875 A.2d 510 (2005) (“our function is not to decide factual issues de novo” (internal quotation marks omitted)).2

Accordingly, I disagree with the majority that the state violated Napue and that the petitioner’s conviction should be reversed on that basis.

The petitioner claims, however, that his convic- tion must nonetheless be reversed because the state suppressed material impeachment evidence concerning Minto, a key state witness, in violation of Brady and Giglio. Specifically, the petitioner argues that the state failed to disclose (1) two recorded interviews conducted by the Manchester Police Department in which Minto repeatedly sought to trade information about the murder of Shamari Jenkins (victim) for favorable treatment in his own robbery case, and (2) Mambrino’s decision not to sign arrest warrants supported by probable cause for Minto’s alleged involvement in armed robberies in West Hartford and Windsor because it was not in the state’s interest to do so based on Minto’s anticipated testimony at the petitioner’s murder trial. The habeas court concluded that the recorded interviews were not material and that the state had no obligation to disclose the decisions not to prosecute Minto for those robberies. I respectfully disagree with the habeas court and, for the reasons set forth herein, agree with the petitioner that his conviction must be reversed.

The majority opinion sets forth the relevant facts and procedural history, as recounted by the Appellate Court in the petitioner’s direct appeal. Accordingly, I do not repeat them in full. Instead, I highlight only those facts that are pertinent to the petitioner’s claim. In 2013, the petitioner and Matthew Allen Hall-Davis, a close friend of his, conspired to kill the victim, who was pregnant with the petitioner’s child and chose not to have an abortion. In April of that year, Hall-Davis approached a vehicle in which the victim and the petitioner were seated and fired a single shot from a Ruger Super Blackhawk .44 Magnum revolver (.44 Ruger) through the rear windshield, fatally striking the victim.

Several days after the murder, Hall-Davis met with Minto, a mutual friend of both Hall-Davis and the petitioner, at Minto’s residence. During that meeting, Hall-Davis confessed that he had killed the victim at the remarks, standing alone, were so egregious as to deprive the petitioner of due process is not before us. petitioner’s behest in order to terminate her pregnancy. Hall-Davis then asked Minto for money to flee the area, but Minto told him that he had none to give him.

On May 11, 2013, Minto and Hall-Davis robbed a jew- elry store in Manchester. During the robbery, Hall-Davis brandished a .44 Ruger. After the petitioner viewed surveillance footage of the robbery on the Internet, he became upset because the footage revealed that Hall- Davis still possessed the same .44 Ruger used to kill the victim, even though he previously told the petitioner that he had disposed of it. The petitioner then contacted Detective Reginald Early of the Hartford Police Depart- ment and reported that Hall-Davis and Minto had com- mitted the Manchester robbery. Based largely on that information, the Manchester Police Department obtained arrest warrants for both men, and Minto surrendered to the police on May 25, 2013. Shortly thereafter, Man- chester police detectives conducted two interviews of Minto. Both interviews were video recorded. During the second interview, Minto, in response to questioning, essentially admitted to participating in two additional robberies in West Hartford and Windsor. Mambrino did not authorize warrants charging Minto with those robberies, expressing that it was not in the state’s inter- est to bring additional charges against him based on his anticipated testimony at the petitioner’s criminal trial. At the petitioner’s criminal trial, the state relied heavily on Minto’s testimony to establish both the peti- tioner’s motive and his participation in the conspiracy with Hall-Davis. Minto testified that Hall-Davis had told him that the petitioner “wanted to get rid of the baby . . . before [the victim] hit seven months” of pregnancy.

Minto further testified that Hall-Davis confessed that he had killed the victim for the petitioner because the petitioner “kept pleading with [him] to do it . . . .” Minto also linked Hall-Davis to the murder weapon, testified that the same firearm had been used in both the homicide and the Manchester robbery, stated that Hall-Davis and the petitioner had repaired a malfunction in the firearm before the murder, and testified that he had observed Hall-Davis dispose of the shell casing from the fatal shot. Because no one other than Hall-Davis and the petitioner witnessed the shooting, Minto’s testimony assumed particular significance.

Following his conviction, the petitioner submitted a Freedom of Information Act (FOIA) request to the Man- chester Police Department seeking records relating to Minto’s cooperation with law enforcement. Through that request, he learned for the first time that the Manchester police had conducted two recorded interviews of Minto that had never been disclosed to the defense. Although the Manchester Police Department had destroyed the recordings pursuant to a court order, the petitioner ultimately obtained copies from the Windsor Police Department.

The first interview reveals that, after initially deny- ing any involvement in the Manchester robbery, Minto quickly shifted to attempting to bargain with the detec- tives once they confronted him with incriminating evi- dence. He stated: “All right. All right. . . . Let’s get it done. . . . What is it going to take for me to get out of here? I got some information.” When the detectives asked what information he possessed, Minto responded: “Murder. . . . Shamari, the East Hartford girl [who] was shot on Mather Street.” Throughout the interview, Minto repeatedly sought leniency in exchange for his cooperation, making statements such as: “Drop all this, I tell you everything,” and “I want to make a deal. If you could help me, I’ll say whatever you want me to say.” The interview further demonstrates that Minto’s will- ingness to cooperate fluctuated depending on whether the detectives suggested that his cooperation might benefit him. After Minto implicated both the petitioner and Hall- Davis in the victim’s murder, the detectives indicated that they already knew much of the information he was providing, at which point Minto resumed denying his involvement in the Manchester robbery and declined to provide additional information. Later, however, after the detective suggested that helping the police recover the firearm could benefit Minto, he immediately resumed negotiating, asking the detectives to “lessen up some of them charges” and repeatedly inquiring, “what are you going to do for me?”

During the second interview, conducted later that same day, Minto identified the .44 Ruger as the murder weapon and sought a reduction in his bond in exchange for assisting the police in recovering it. He began the interview by stating: “You get the bond. I’m going to get you the weapon.” The detective responded that recover- ing the firearm would be “a huge thing” for Minto and assured him that the prosecutor would be informed of his cooperation, while cautioning that he could make no promises because of the seriousness of the charges. The second interview also revealed Minto’s involvement in additional robberies in West Hartford and Windsor. The detective informed Minto that those jurisdictions possessed sufficient evidence to arrest him and advised him not to deny his involvement because “the whole deal falls through” if he was not truthful about the addi- tional robberies. The detective further explained that Manchester, West Hartford, and Windsor had formed a joint task force to investigate those crimes. Following that discussion, Minto effectively admitted his participa- tion by confirming details of both robberies in response to the detective’s questions.

The petitioner subsequently submitted additional FOIA requests to the West Hartford and Windsor Police Departments. The records produced in response estab- lished that both departments had prepared arrest war- rants to charge Minto in connection with separate armed robberies and that Mambrino, the prosecutor who was also assigned to the petitioner’s murder, declined to authorize those warrants.

The petitioner thereafter filed an amended petition for a writ of habeas corpus. In his operative amended peti- tion, he alleged, among other claims, that the prosecuting authority violated his rights to due process and a fair trial by failing to disclose material exculpatory evidence relating to Minto. At the habeas trial, the petitioner introduced the Manchester interview videos and evidence that Mambrino had declined to issue arrest warrants for the West Hartford and Windsor armed robberies. Mam- brino testified that her declination of the West Hartford warrant had nothing to do with Minto’s cooperation. She explained that both the West Hartford and Windsor war- rants had been denied in the interest of judicial economy and not because Minto had agreed to cooperate with the state. Following the trial, the habeas court denied the petitioner’s petition.

With respect to the Manchester interview videos, the habeas court concluded that they were not material evi- dence. The court stated that “[t]he fact that Minto had pending cases in Manchester was known to the defense and he was questioned about his motivation for testify- ing. [Although] it certainly would have been better for [the petitioner’s trial counsel] to confront Minto about his requests for consideration during that interview, the court cannot conclude that there is a reasonable likelihood that those questions would have altered the jury’s ver- dict.” The habeas court likewise rejected the petitioner’s claim concerning the state’s failure to disclose its decision not to prosecute the West Hartford and Windsor robber- ies, explaining that “[t]here is no evidence that Minto was aware that those charges were not being prosecuted in exchange for his testimony, or even that [Mambrino] did it for that purpose. The evidence establishe[d] that [Mambrino] unilaterally decided not to prosecute those charges because she believed there were already enough charges to prosecute Minto on.” Following the denial of the petition, the habeas court granted the petitioner’s petition for certification to appeal, and the petitioner appealed to the Appellate Court.

After filing his principal brief in the Appellate Court, the petitioner sought permission from the Appellate Court to file a motion for augmentation and rectification of the record with the habeas court. In support of that motion, he represented that, during the pendency of the appeal, he had discovered, through an additional FOIA request, an email exchange between Mambrino and a West Hartford police detective. According to the petitioner, the email exchange established, contrary to Mambrino’s sworn testimony at the habeas trial, that she had declined to authorize arrest warrants charging Minto with multiple armed robberies because it was “not in [the state’s] best interest” to prosecute him in light of his anticipated testimony at the petitioner’s criminal trial. The Appellate Court granted that motion and directed the appellate clerk “to forward the motion for rectification and response to the habeas court for consideration.”3

The habeas court held a hearing on the motion, at which the parties agreed that the newly discovered evi- dence—specifically, the email from Mambrino to the police detective—would be admitted as a full exhibit and made a part of the record. Following the hearing, the habeas court concluded that “the existence of the email sent by [Mambrino] does not alter [the court’s] application of the prevailing case law.” In explaining that conclusion, the court observed: “What the email proves is that [Mambrino] declined to sign warrants charging Minto with additional crimes because he was going to testify against [the petitioner], and therefore it was not in the state’s best interest to arrest him. There is a fair argument to be made that this information constitutes information favorable to the defense that must be dis- closed.” The habeas court nevertheless concluded that the newly discovered evidence did not affect its resolu- tion of the petitioner’s claim because, in the operative habeas petition, he did not allege “that the state [had] failed to disclose exculpatory or impeachment infor- mation in the general sense . . . .” Rather, the court explained, the petitioner alleged the narrower claim that the state “failed to disclose that Minto [had] received favorable [treatment] in exchange for his cooperation . . . .” (Emphasis omitted.) Accordingly, because there remained “simply no evidence of an agreement between the state and Minto or that he was aware that the state had chosen to decline warrants for his involvement in other offenses,” the court determined that the newly discovered email did not alter its analysis.

The petitioner thereafter filed in the Appellate Court a motion seeking a remand to the habeas court for recon- sideration of its ruling on the motion for rectification. On the same day, the petitioner filed what he characterized as an “alternative motion,” renewing his request for rec- tification and seeking a remand to permit him to amend his habeas petition. The Appellate Court denied both motions but ordered the petitioner to file a supplemental brief addressing the issues raised in the two motions. After briefing in the Appellate Court concluded, but before oral argument was scheduled, the petitioner filed a motion pursuant to Practice Book § 65-2 requesting that this court transfer the appeal to itself. We granted that motion.

As a threshold matter, the petitioner claims that the habeas court abused its discretion during the rectifica- tion proceedings by declining to consider whether the state’s suppression of the newly discovered evidence— namely, the email exchange between Mambrino and a West Hartford detective—deprived him of a fair trial. The petitioner contends that the habeas court improperly construed his operative petition as alleging only that the state had failed to disclose an agreement between the state and Minto, pursuant to which Minto received favorable treatment in exchange for his cooperation, rather than as a broader claim that the state had violated its obligations under Brady by suppressing material impeachment evidence generally. I agree with the peti- tioner. “[T]he interpretation of pleadings is always a question of law for the court . . . .” (Internal quotation marks omitted.) Carpenter v. Commissioner of Correction, 274 Conn. 834, 842, 878 A.2d 1088 (2005). In interpreting pleadings, Connecticut courts construe them “broadly and realistically, rather than narrowly and technically.” (Internal quotation marks omitted.) Oliphant v. Com- missioner of Correction, 274 Conn. 563, 569, 877 A.2d 761 (2005). “[W]e long have eschewed the notion that pleadings should be read in a hypertechnical manner.” (Internal quotation marks omitted.) Flannery v. Singer Asset Finance Co., LLC, 312 Conn. 286, 299, 94 A.3d 553 (2014). “Although essential allegations may not be supplied by conjecture or remote implication . . . [a habeas petition] must be read in its entirety in such a way as to give effect to the pleading with reference to the general theory [on] which it proceeded, and do substantial justice between the parties. . . . As long as the pleadings provide sufficient notice of the facts claimed and the issues to be tried and do not surprise or prejudice the opposing party, we will not conclude that the [habeas petition] is insuf- ficient to allow recovery.” (Footnote omitted; internal quotation marks omitted.) Carpenter v. Commissioner of Correction, supra, 842.

In count two of the operative habeas petition, the peti- tioner alleged a constitutional due process violation based on “the prosecuting authority’s failure to disclose material exculpatory evidence.” The petitioner specifi- cally alleged that “[t]he prosecuting authority failed to disclose in a timely manner the following evidence to the petitioner or his counsel prior to the petitioner’s criminal trial . . . (B) [that Minto] received favorable disposi- tions in [State v. Minto, Superior Court, judicial dis- trict of Hartford, Docket No. HHD-CR-XX-XXXXXXX-T]; Manchester Police Department Case No. M201304348; Windsor Police Department Case No. 2013-21679; and West Hartford Police Department Case No. 1300023062, Incident No. 083, in exchange for his cooperation with the police and/or prosecuting authority in [State v. Bryan, Superior Court, judicial district of Hartford, Docket No. HHD-CR-XX-XXXXXXX-T] . . . .” In his habeas petition, the petitioner further described each of these items as “excul- patory or otherwise favorable evidence that should have been disclosed to the petitioner or his counsel prior to his criminal trial” and alleged that “there is a reasonable probability that, had [they] been disclosed in adequate time to be used by the defense during the petitioner’s criminal trial, the result of the petitioner’s criminal trial would have been different and more favorable to the petitioner.”

I conclude that the habeas court construed the opera- tive habeas petition too narrowly. In it, the petitioner alleged a general claim under Brady and then specified that the prosecuting authority failed to disclose that Minto had “received favorable dispositions” in multiple criminal matters “in exchange for his cooperation with the police and/or the prosecuting authority . . . .” Prop- erly understood, that allegation is not limited to the existence of a formal or informal agreement between the state and Minto. Rather, read broadly and realistically, as our pleading jurisprudence requires, the petitioner alleged that the state had failed to disclose favorable impeachment evidence bearing on Minto’s motive to cooperate with law enforcement and to testify against the petitioner. Such an allegation falls squarely within the general principles announced in Brady, and it pro- vided the respondent, the Commissioner of Correction, with fair notice that the petitioner claimed, in general terms, a violation of the state’s disclosure obligations under Brady. Having concluded that the petitioner’s claim was sufficiently pleaded and that the record is adequate for review, I turn to the merits of his Brady claim, namely, whether the state’s suppression of (1) its pretrial decision not to authorize arrest warrants, despite having probable cause, for Minto’s alleged armed robberies in Windsor and West Hartford, and (2) video recordings of Minto’s interviews with the Manchester Police Department constituted violations of Brady.4 C Under Brady, the government bears a constitutional duty to disclose favorable evidence to the accused when such evidence “is material either to guilt or to punish- ment . . . .” Brady v. Maryland, supra, 373 U.S. 87. Favorable evidence includes not only evidence that tends to exculpate the accused, but also evidence that is use- ful to impeach the credibility of a government witness. See Giglio v. United States, supra, 405 U.S. 153–54; see also Greene v. Commissioner of Correction, supra, 330 Conn. 15.

The rules established in Brady and its progeny derive from the due process clauses of the United States consti- tution, and their core aim is to protect a defendant’s right to a fair trial by ensuring the reliability of any criminal verdict against him. See United States v. Bagley, supra, 473 U.S. 675. In order to prove a Brady violation, “a defendant must show that (1) the evidence is favorable to the accused, either because it is exculpatory, or because it is impeaching, (2) the evidence was suppressed by the state either wilfully or inadvertently, and (3) the evidence was material to the case.” (Internal quotation marks omitted.) Vega v. Commissioner of Correction, 354 Conn. 437, 445, 354 A.3d 151 (2026); see also Strickler v. Greene, 527 U.S. 263, 281–82, 119 S. Ct. 1936, 144 L. Ed. 2d 286 (1999). Evidence is material “only if there would be a reasonable probability of a different result if the evidence had been disclosed.” (Internal quotation marks omitted.) State v. Jordan, 314 Conn. 354, 370, 102 A.3d 1 (2014); see also Kyles v. Whitley, 514 U.S. 419, 433–34, 115 S. Ct. 1555, 131 L. Ed. 2d 490 (1995). The petitioner first claims that “the habeas court erred in finding that the state did not have an obligation to disclose its decision not to prosecute [Minto]” in con- nection with the Windsor and West Hartford robberies. He argues that the authorities on which it relied were inapposite and that it incorrectly determined that the arrest warrant information did not have to be disclosed on the basis that there was no evidence Minto was aware that the charges were not being prosecuted in exchange for his testimony. In the petitioner’s view, the habeas court’s “constrained application of Brady,” which would “prohibit a finding that this information was favorable . . . unless it establishe[d] an incentive promised to the witness, is wholly inconsistent with Brady’s require- ment that the prosecutor disclose all favorable exculpa- tory impeachment information.” (Emphasis omitted.)

The respondent counters that the habeas court did not err because the petitioner had failed to demonstrate that the warrants were denied pursuant to “an agree- ment or understanding, whether formal or informal, in exchange for Minto’s testimony at the petitioner’s criminal trial.” I agree with the petitioner that, because Mambrino declined the warrants based on Minto’s antici- pated cooperation, the state conferred a benefit on Minto that it was obligated to disclose to the defense.

The habeas court’s reasoning—and the respondent’s argument on appeal—rests on the premise that favor- able treatment or benefits afforded to a cooperating witness do not qualify as impeachment evidence unless a defendant first establishes the existence of an express or implied agreement between the witness and the pros- ecution. That premise misconstrues the nature of the state’s disclosure obligations under Brady and Giglio. The state’s duty to disclose impeachment evidence is not limited to circumstances in which there is proof of an informal or formal agreement. Rather, impeachment evidence is “broadly defined” as “evidence having the potential to alter the jury’s assessment of the credibility of a significant prosecution witness.” (Internal quotation marks omitted.) Adams v. Commissioner of Correction, 309 Conn. 359, 369–70, 71 A.3d 512 (2013).

Here, the evidence demonstrates that Mambrino declined to authorize warrants supported by probable cause for the arrest of Minto for two armed robberies because pursuing those charges was not in the state’s interest in light of his anticipated testimony against the petitioner. By withholding authorization for the warrants, and by failing to disclose that decision and the basis for it, the state deprived the petitioner of favor- able impeachment evidence. This evidence would have tended to impeach Minto’s credibility by demonstrating a motive to testify favorably for the state by implicating the petitioner in the victim’s murder.5 See, e.g., Benn v. Lambert, 283 F.3d 1040, 1057 (9th Cir.) (explaining that disclosure of benefits conferred on informant “could have allowed the jury to reasonably conclude that [the informant] had a motive other than altruism for testify- ing on behalf of the [s]tate” (internal quotation marks omitted)), cert. denied, 537 U.S. 942, 123 S. Ct. 341, 154 L. Ed. 2d 249 (2002); United States v. Johnson, 26 F.3d 669, 682–83 (7th Cir.) (expressing concern over govern- ment’s failure to disclose, under Brady and Giglio, that it had requested state prosecutor to delay filing robbery charges against witness until after she testified), cert. denied, 513 U.S. 940, 115 S. Ct. 344, 130 L. Ed. 2d 300 (1994). Although the respondent contends that the evidence lacked impeachment value because the habeas court found no evidence that Minto knew that Mambrino had declined to pursue those charges in exchange for his testimony, evidence of the witness’ awareness of the benefit is not a prerequisite for the evidence to possess impeachment value. See, e.g., State v. Jordan, supra, 314 Conn. 370 (explaining that impeachment evidence is “broadly defined” (internal quotation marks omitted)). A benefit conferred on a cooperating witness, regardless of specific evidence that he knew of that benefit, can bear on his credibility because it is objective evidence of the relationship between the witness and the state, evidence the jury is entitled to weigh in assessing whether his testimony was the product of an ordinary civic obliga- tion or something else.6

Turning to the Manchester interview videos, I simi- larly conclude that they were favorable evidence that should have been disclosed to the defense. During those interviews, Minto repeatedly attempted to negotiate with law enforcement for favorable treatment in exchange for providing information concerning the victim’s death. Indeed, at one point, Minto stated: “I want to make a deal. If you could help me, I’ll say whatever you want me to say.” That statement alone was powerful impeachment evidence because it demonstrated Minto’s willingness to tailor his account if doing so would secure a benefit from the state. A reasonable juror could have viewed the statement as reflecting not merely a generalized desire for leniency but a willingness to subordinate the truth to his own self-interest.

Additionally, during the interviews, law enforcement informed Minto that he was a suspect in two additional armed robberies and that the investigating agencies possessed sufficient evidence to obtain arrest warrants against him. Thus, at the very time Minto was attempt- ing to negotiate with investigators, he knew that he faced the realistic prospect of prosecution for multiple, additional serious felonies. That circumstance provided a concrete and compelling reason for him to seek favor- able treatment from the state and, in turn, to provide information that would be perceived as valuable to the prosecution. The videos of the interviews would have permitted the defense to argue that Minto was motivated by his own substantial criminal exposure and was actively attempting to leverage testimony against the petitioner into favorable treatment for himself. Accordingly, I conclude that the suppressed evidence was favorable to the defense.

The remaining question is whether the suppressed impeachment evidence was material. Evidence is mate- rial “if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.” (Internal quota- tion marks omitted.) Kyles v. Whitley, supra, 514 U.S. 433–34. Put differently, “[t]he question is not whether the defendant would more likely than not have received a different verdict with the evidence, but whether in its absence he received a fair trial, understood as a trial resulting in a verdict worthy of confidence. A reasonable probability of a different result is accordingly shown when the government’s evidentiary suppression under- mines confidence in the outcome of the trial.” (Internal quotation marks omitted.) Adams v. Commissioner of Correction, supra, 309 Conn. 370–71, quoting Kyles v. Whitley, supra, 434.

In the present case, there can be little dispute that Minto occupied a central role in the state’s case. First, through his testimony, the state introduced Hall-Davis’ admissions, which directly implicated the petitioner in the victim’s death. Minto testified that Hall-Davis had confessed to him that he killed the victim and did so at the petitioner’s behest. Minto’s testimony painted the petitioner as the architect of the murder, one who requested Hall-Davis to carry it out.

Second, Minto linked the petitioner to the murder weapon. Minto testified that, some time before the vic- tim’s murder, he saw the petitioner and Hall-Davis with a .44 Ruger and that the two went into another room of the apartment they were in to repair it by installing a spring. Minto testified that the .44 Ruger was the same firearm Hall-Davis had used during their Manchester robbery and to kill the victim. He indicated that, after the robbery, Hall-Davis discarded a shell casing and told him it was from the bullet that killed the victim. Minto further testified that he then drove to Henry Park in Vernon, where he dropped off Hall-Davis so that Hall- Davis could hide the gun. Minto testified that, after the robbery, he, Hall-Davis, and the petitioner met at a restaurant, where the petitioner was upset with Hall- Davis because he had not discarded the firearm and said that he wanted the gun back. Notably, without Minto, law enforcement would not have recovered the firearm. After his arrest by the police for the Manchester robbery, Minto led the police to the area of Henry Park where he had dropped off Hall-Davis to hide the gun.

Third, Minto was instrumental in advancing the state’s theory of motive. Minto testified that, in January 2013, the petitioner told him the victim was pregnant with his child and that he was trying to convince her to have an abortion. Minto later testified that, when Hall-Davis admitted killing the victim, Hall-Davis said the petitioner wanted her dead in order to terminate the pregnancy before she reached seven months. Consequently, Minto’s credibility was of particular importance to the jury’s assessment of the state’s case.

To the extent there was any ambiguity as to Minto’s importance to the state’s case, the degree to which Mam- brino relied on Minto’s testimony in closing argument clears it up. In closing, Mambrino identified the elements the state needed to prove and directed the jury to Minto’s testimony as evidence supporting each element. As to the murder count, Mambrino argued, “[i]n count one, the state has to prove two elements: the [petitioner] intended to cause the death of another person, and the [petitioner] caused the death of such person. Evidence to support element one: Look at [Minto’s] testimony. He told you that [the petitioner] was upset with [the victim] when she told him that she was going to have the baby. He did not want to have a baby with her.” As to causation, Mambrino argued, “Hall-Davis also told [Minto] that he’s the one [who] shot and killed [the victim] because [the petitioner] pressured him to do it.” Mambrino again emphasized Minto’s testimony in arguing that the peti- tioner was guilty as an accessory to murder, stating, the petitioner “is charged as an accessory in this case. And the judge is going to tell you that, in order for the state [to] prove that someone is an accessory . . . he either solicited or requested or commanded or importuned or intention- ally aided [Hall-Davis] to kill [the victim]. [Hall-Davis] told [Minto] that the [petitioner] begged him to do it. . . . [The petitioner] then aided [Hall-Davis] by providing him with his gun . . . and . . . when [Reginald Lewis, one of the petitioner’s former coworkers] couldn’t fix that gun, both Hall-Davis and [the petitioner] fixed that gun.” The respondent contends that the interview recordings and the pretrial decision not to authorize arrest war- rants for Minto’s alleged armed robberies in Windsor and West Hartford were not material evidence. With respect to the decision not to authorize the arrest war- rants, the respondent maintains that such evidence has no impeachment value. As to the interview recordings, the respondent contends that the jury already knew that Minto had cooperated with law enforcement in the Manchester robbery investigation, understood that he expected to receive some benefit in exchange for that cooperation, and heard extensive cross-examination on those subjects. In essence, the respondent’s position is that the withheld evidence was immaterial because it was merely cumulative impeachment evidence or because defense counsel had already impeached Minto’s cred- ibility through other means.

Although Minto was subjected to cross-examination and the jury was aware that Cohen had offered Minto’s cooperation to the state in exchange for making Minto’s cooperation known to the sentencing judge, the sup- pressed evidence was qualitatively different from the impeachment evidence presented at the petitioner’s criminal trial. See, e.g., Turner v. United States, 582 U.S. 313, 327, 137 S. Ct. 1885, 198 L. Ed. 2d 443 (2017) (explaining that impeachment evidence is not immaterial merely because witness may have already been impeached with other evidence). As previously discussed, during the recorded interviews, Minto repeatedly sought leniency in exchange for his cooperation. Perhaps most damning was when he said to detectives: “I want to make a deal. If you could help me, I’ll say whatever you want me to say.” (Emphasis added.) This statement was direct evidence, in Minto’s own words, that he was willing to tailor his account to secure a benefit from the state. Such evidence strikes at the heart of a witness’ credibility in a manner fundamentally different from generalized evidence of bias or cooperation.

The withheld evidence also would have revealed that Minto faced substantially greater criminal exposure than the jury was led to believe. That information, together with the fact that law enforcement had made the tactical decision not to authorize warrants because of Minto’s cooperation in the petitioner’s case, would have furnished defense counsel with further ammunition to argue that Minto’s testimony was motivated by self-interest rather than truthfulness.

The respondent nonetheless argues that there is not a reasonable probability that a different result would have occurred because the state’s case against the petitioner was strong. The materiality inquiry under Brady, how- ever, “is not a sufficiency of [the] evidence test.” Kyles v. Whitley, supra, 514 U.S. 434. The relevant question is not whether the remaining evidence, viewed in isolation, would have been sufficient to sustain the conviction but whether the suppressed evidence, considered collectively, “could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict.” Id., 435. On the facts of this case, and given Minto’s central role in the prosecution, I conclude that it could. The withheld evidence would have provided the defense with a substantially stronger basis on which to challenge the credibility of Minto, the state’s principal witness, whose testimony was central to the state’s theory that the petitioner had orchestrated the murder. In sum, I am persuaded that the cumulative effect of the suppressed impeachment evidence sufficiently undermines confi- dence in the verdict.

Because I would reverse the judgment of the habeas court and remand the case with direction to grant the habeas petition, to vacate the petitioner’s underlying conviction, and to order a new trial on the basis of the foregoing Brady violations, I concur in the judgment.

1 This court’s decision in Greene v. Commissioner of Correction, supra, 330 Conn. 1, bears a striking resemblance to the present one. As in the present case, the petitioner in Greene claimed that the state had vio- lated Napue by failing to correct testimony from a cooperating witness concerning the nature of the witness’ agreement with the state. Id., 9, 13. This court rejected that claim, emphasizing that the challenged testimony must be evaluated in context; id., 21–22; and concluding that the habeas court reasonably found that the testimony, although “ ‘not a model of clarity,’ ” was “neither false nor substantially mislead- ing.” Id., 24. 2 Because I conclude that there was no Napue violation, I do not agree with the majority’s conclusion that Mambrino bolstered false or sub- stantially misleading testimony during her rebuttal closing argument. My disagreement on that point, however, should not be understood to suggest that I condone Mambrino’s characterization of the understand- ing between the state and Minto when she argued that Minto did not “have an interest in the outcome of this case.” The issue of whether these 3 On the same day that the petitioner moved for permission to file a motion for augmentation and rectification with the habeas court, he also filed a motion to remand for further habeas proceedings. The Appellate Court indicated that no action on that motion was necessary. 4 There is no dispute that the petitioner never received or was aware of this information and, therefore, that the information was suppressed. Notably, on February 25, 2015, before his criminal trial, the petitioner filed a “Motion for Giglio Material,” in which he requested a host of information from the state. He asked for, among other things, “[a] complete list of all crimes committed by the informant(s) and witness(es) where there was no arrest or apprehension, including dates and places,” “[t]he specific consideration received or promised in exchange for the informant(s)’ and witness(es)’ services to the state, including but not limited to . . . [f]avorable disposition of any criminal or civil cases,” and “[d]isclosure of any and all files, confidential or otherwise, relating to the informant(s) and witness(es). These files may exist with respect to the [Federal Bureau of Investigation], the [Drug Enforcement Administra- tion], state and local police agencies, prosecutor’s offices, or organized crime task force[s].” 5 A prosecutor undoubtedly retains broad discretion to decline to authorize an arrest warrant or to file a criminal information. When that discretion is exercised in good faith because the prosecutor concludes that probable cause is lacking, or because other legitimate, nontactical reasons exist, the decision may not constitute a conveyed benefit. When, however, as in the present case, the decision not to authorize charges is a trial tactic—i.e., to minimize impeachment evidence against a state’s witness—the state has given that witness a benefit that must be disclosed under Brady and Giglio. 6 In concluding that that Mambrino’s decision not to authorize the warrants for the Windsor and West Hartford robberies was not informa- tion that needed to be disclosed to the defense, the habeas court relied on the Second Circuit’s decision in Shabazz v. Artuz, 336 F.3d 154, 165 (2d Cir. 2003). Shabazz, however, is inapposite to the present case because it addressed a fundamentally different question. The issue in Shabazz was whether the petitioner had established the existence of an alleged, undisclosed promise of leniency between the prosecution and its cooperating witnesses. Id., 163–65. The Second Circuit held only that the prosecutor’s decision to afford favorable treatment to a wit- ness, standing alone, does not permit an inference that the prosecutor had made a promise of leniency in exchange for the witness’ testimony. Id., 165. The court explained that “[t]he government is free to reward witnesses for their cooperation with favorable treatment in pending criminal cases without disclosing to the defendant its intention to do so, provided that it does not promise anything to the witnesses prior to their testimony.” (Emphasis in original.) Id. Here, however, the issue is not whether the state made an undisclosed promise of leniency. Rather, the question is whether Mambrino’s decision to confer a substantial pretrial benefit on an important cooperating witness—one made for the tactical purpose of minimizing the impeachment evidence available against that witness—constituted favorable impeachment evidence that the state was required to disclose to the defense.

Opinion of the Court

OF CORRECTION (SC 21193)

Mullins, C. J., and McDonald, D’Auria, Ecker, Dannehy and Bright, Js.

Syllabus The petitioner, who had been convicted of murder and conspiracy to commit murder in connection with the shooting death of the victim, sought a writ of habeas corpus. The petitioner had plotted with H, the petitioner’s close friend, to kill the victim, who was pregnant with the petitioner’s child. The petitioner allegedly sought to terminate the victim’s pregnancy after he failed to convince her to have an abortion. H ultimately fired the fatal gunshot through the rear window of a car that the victim had been driving. A few weeks after the victim’s murder, H met with M, a mutual friend, and H confessed that he had murdered the victim at the petitioner’s behest. During the investigation of a robbery that H and M had committed together after the victim’s murder, the police gathered information implicating H and the petitioner in the murder. At the petitioner’s criminal trial, M testified for the prosecution, stating, inter alia, that H had confessed to committing the murder at the petitioner’s behest and that H had used the same gun during both the victim’s murder and the robbery that H and M had committed together. During his testimony, M denied that he had a deal or agreement with the state or that anything was promised to him relating to pending criminal charges against him in consideration for his testimony against the petitioner. M, however, stated during recross-examination that he had “considerations, but those are the details that will happen,” without specifying what those details were. The prosecution also called M’s defense attorney, C, to testify at the petitioner’s criminal trial, and C testified that there was no agreement between M and the state but that there “would be some consideration given” to M if he testified truthfully at the petitioner’s criminal trial. In the habeas court, the petitioner claimed, inter alia, that the prosecutor had violated Napue v. Illinois (360 U.S. 264) when she knowingly presented M’s and C’s false or substantially misleading testimony regarding whether M had a deal or agreement with the state to exchange his testimony against the petitioner for consideration in M’s own criminal case. The habeas court rendered judgment denying the habeas petition, and the petitioner, on the granting of certification, appealed. Held: The prosecutor violated Napue by eliciting false or substantially misleading testimony from M and C concerning M’s agreement with the state, failing to correct the false impressions created by their testimony, and adopting or relying on that testimony during closing argument, and, because those Napue violations were material, this court reversed the habeas court’s judgment and remanded the case with direction to grant the habeas petition, to vacate the petitioner’s conviction, and to order a new criminal trial.

The habeas court’s findings that M did not testify falsely insofar as there merely was some ambiguity in his testimony due to the “different terms” that he and the prosecutor used to describe the arrangement M had with the state, and that this ambiguity was later clarified were clearly erroneous in view of M’s blanket denial of the existence of an agreement, which was never corrected by the prosecutor.

Moreover, on redirect examination, the prosecutor reinforced the idea that M had no incentive to wrongly implicate the petitioner by having M confirm that he had no expectation that anything would be given to him as a result of his testimony against the petitioner and that he was testifying because it was “the right thing to do,” and the habeas court’s reliance on M’s statement on recross-examination that he had “considerations, but those are the details that will happen,” to support that court’s conclusion that the jury was made aware of M’s arrangement with the state was misplaced.

Furthermore, C’s testimony that there was no agreement but that there would be some consideration if M testified truthfully at the petitioner’s criminal trial did not elucidate the nature of the actual agreement that existed in this case, and C’s testimony that there was no consideration for M other than merely “the hope, perhaps, the expectation” that the sentencing court in M’s criminal case would consider M’s cooperation did not put the jury in the petitioner’s criminal case on notice that there had always been an agreement between M and the state, and, therefore, contrary to the habeas court’s conclusion, C’s testimony was substantially misleading.

In addition, the habeas court’s finding that the prosecutor sufficiently clari- fied the nature of the state’s agreement with M was clearly erroneous in light of how the prosecutor viewed the totality of the testimony and described it to the jury during her closing argument.

Specifically, the prosecutor argued to the jury that M had no interest in the outcome of the petitioner’s case, despite being fully aware that M’s outstand- ing criminal charges would not be resolved until after the outcome of the petitioner’s criminal trial and his sentence were known, and the prosecutor failed to inform the jury during closing argument that there was, at a mini- mum, an agreement to bring M’s cooperation with the state to the attention of the sentencing court in M’s criminal case but that there was no agreement as to a specific sentence.

The Napue violations were material, as there was a reasonable likelihood that M’s and C’s false or substantially misleading testimony concerning the lack of an agreement between M and the state could have affected the jury’s verdict.

M’s testimony directly implicated the petitioner in the victim’s murder and was the only evidence linking the murder weapon to the crimes charged, the Napue violations enhanced M’s credibility, and, after accounting for the Napue violations, this court was not persuaded that the remainder of the state’s case against the petitioner was strong enough to render the Napue violations harmless.

(Two justices concurring in the judgment and one justice dissenting in part in one separate opinion) Argued April 15—officially released September 1, 2026 Procedural History Amended petition for a writ of habeas corpus, brought to the Superior Court in the judicial district of Tolland, geographical area number nineteen, and tried to the court, Bhatt, J.; judgment denying the petition, from which the petitioner, on the granting of certification, appealed. Reversed; judgment directed.

Kayla R. Stephen, deputy assistant public defender, for the appellant (petitioner).

Meryl R. Gersz, assistant state’s attorney, with whom, on the brief, were Sharmese L. Walcott, state’s attorney, and Christopher Alexy, senior assistant state’s attorney, for the appellee (respondent).

Opinion

McDONALD, J. The petitioner, Carlton Bryan, appeals from the judgment of the habeas court, which denied his third amended petition for a writ of habeas corpus chal- lenging his conviction of murder in violation of General Statutes §§ 53a-54a (a) and 53a-8, and conspiracy to commit murder in violation of § 53a-54a (a) and General Statutes § 53a-48 (a). See State v. Bryan, 193 Conn. App. 285, 287–88, 293–94, 219 A.3d 477, cert. denied, 334 Conn. 906, 220 A.3d 37 (2019). He claims that the state violated Napue v. Illinois, 360 U.S. 264, 269, 79 S. Ct. 1173, 3 L. Ed. 2d 1217 (1959), when the prosecutor, Senior Assistant State’s Attorney Donna Mambrino, knowingly presented the false or substantially mislead- ing testimony of two state witnesses, Kingsley Minto and Richard Cohen, Minto’s defense counsel, during the petitioner’s criminal trial regarding whether Minto made a “deal” with the state to exchange his testimony for consideration in his own criminal case. Additionally, the petitioner claims that Mambrino’s endorsement of this false or substantially misleading testimony in her closing argument contributed to the Napue violations. Because we agree that the state violated Napue, we reverse the judgment of the habeas court.1

The Appellate Court’s decision in the petitioner’s direct appeal, as supplemented by the decision of the habeas court, sets forth the relevant facts that reasonably could have been found by the jury, and the procedural history, which we summarize. See State v. Bryan, supra, 193 Conn. App. 288–94. The victim, Shamari Jenkins, was pregnant with the petitioner’s child. Having failed to convince the victim to have an abortion, the petitioner plotted with a close friend, Matthew Allen Hall-Davis, to kill the victim in order to terminate the pregnancy. In order to carry out the plan, the petitioner asked a for- mer coworker, Reginald Lewis, to repair a Ruger Super Blackhawk .44 Magnum revolver (.44 Ruger). Lewis was unable to fix the .44 Ruger, but Hall-Davis told Lewis that he would fix it. Sometime later, the petitioner and Hall-Davis repaired the .44 Ruger.

Prior to the shooting, the petitioner, the victim, and Hall-Davis attended a cookout at the home of the vic- tim’s father in East Hartford. The petitioner and Hall- Davis left the cookout together and met with Hall-Davis’ cousin, Everett Walker, near Walker’s apartment build- ing, which was located on Magnolia Street in Hartford. The petitioner told Walker that the victim refused to have an abortion and that he wanted to kill her in the vicinity of Walker’s apartment building. He also asked Walker to act as a lookout and to tell any police offi- cers dispatched to the crime scene that he had observed an unknown individual running away from the scene.

Walker did not respond to the request and returned to his apartment.

Later, the victim drove the petitioner in her car to Mag- nolia Street, where the petitioner purportedly intended to meet with his cousin. Hall-Davis approached the car and, using the .44 Ruger, fired a single gunshot through the rear window of the car, fatally striking the victim. The petitioner then called 911 to report that the victim had been shot, but he did not identify the shooter.

Subsequently, the petitioner met with Detective Regi- nald Early at the Hartford Police Department and sub- mitted a signed, sworn statement regarding the victim’s murder. In that statement, the petitioner asserted that the shooter was Kevan Simmons. Following an investi- gation, Detective Early ruled out Simmons as a suspect. A few weeks after the murder, Hall-Davis met with Minto, a mutual friend of both Hall-Davis and the peti- tioner. Hall-Davis confessed to Minto that he had killed the victim at the petitioner’s behest because she would not terminate her pregnancy. Hall-Davis then asked Minto for money so that he could flee the area, but Minto had no money to give to Hall-Davis.

Thereafter, Hall-Davis and Minto robbed a jewelry store in Manchester. During the robbery, which was recorded on surveillance video, Hall-Davis brandished the .44 Ruger. As Hall-Davis and Minto were driving away from the jewelry store, Hall-Davis tossed a shell casing out of the car window, which Hall-Davis told Minto was from the bullet that he had fired at the victim. Later that day, they drove to a park in Vernon, where Hall-Davis hid the .44 Ruger.

After viewing the video footage of the Manchester robbery on the Internet, the petitioner became upset that Hall-Davis had lied to him about having disposed of the .44 Ruger. He called Detective Early to report that Hall-Davis and Minto had committed the Manchester robbery. Largely on the basis of that information, the Manchester Police Department secured arrest warrants for Hall-Davis and Minto. While in police custody, Minto admitted to his involvement in the Manchester robbery and assisted the police in locating the .44 Ruger.

On the basis of information that he obtained during the course of his investigation from, among others, Hall- Davis, Minto, and Lewis, Detective Early secured arrest warrants for the petitioner and Hall-Davis in relation to the victim’s murder. After his arrest, the petitioner agreed to be interviewed by Detective Early and another detective. He submitted a signed, sworn statement, alleging that, while he was sitting with the victim in her car on Magnolia Street, Hall-Davis entered the car and sat in the backseat, behind the victim. Detective Early questioned the petitioner as to how Hall-Davis could have entered the car, which had only two doors, without the petitioner first exiting the car. Detective Early also noted that the bullet that struck the victim had been shot through the rear window of the car and would have hit Hall-Davis had he been seated in the backseat. The petitioner ended the interview at that juncture.

At the petitioner’s criminal trial, Mambrino called Minto as a prosecution witness. Through his testimony, which was based in large part on Hall-Davis’ hearsay statements, the prosecution established the petitioner’s motive for the victim’s murder and the petitioner’s con- spiracy with Hall-Davis to commit the murder. Similarly, it was through Minto’s testimony that the prosecution introduced Hall-Davis’ confession to killing the victim at the petitioner’s behest. Minto also testified that Hall- Davis had used the same gun in both the Hartford murder and in the Manchester robbery, that the petitioner and Hall-Davis had fixed a mechanical issue affecting this gun before the murder, that he saw Hall-Davis dispose of a shell casing from the bullet that had been fired at the victim, and that the video footage of the Manchester robbery had upset the petitioner because he saw that Hall-Davis had not discarded the gun.

Given the importance of Minto’s testimony, both Mam- brino and the petitioner’s defense counsel questioned Minto about his potential motives for testifying. On direct examination, the following exchange occurred between Mambrino and Minto:

“Q. Have you been promised any deals in exchange for your testimony?

“A. No.

“Q. Have you been told that, if you testify in this case . . . at the time that you enter a plea in your case . . . the [prosecutor] would merely tell the judge that you cooper- ated and testified in a criminal case?

“A. No.

“Q. Let me rephrase the question. Were you told that, if you testified in this case . . . [when] you [plead] in front of the judge [presiding over your criminal case] . . . the [prosecutor], either me or another member of my office, would tell the judge that you cooperated . . . in [this] case? “A. No deal was made.

“Q. No specific deal was made, but was that told to you? “A. Yes.”

On cross-examination, the following exchange took place between defense counsel and Minto: “Q. . . . You testified on direct that ‘there’s been no deals made.’ [Those are] the words you used, right?

“A. No deals.

“Q. So . . . you’re telling this jury, and you’re testify- ing here, that you have no expectation of anything being given to you for your testimony?

“A. Nothing hasn’t been given to me. Nothing had been promised to me.

“Q. Okay. So, you’re doing it to be like a good citizen. Is that it?

“A. Not in those words.

“Q. Okay. Isn’t it true that you anticipate that . . . at some point in the future, when your robbery case is resolved . . . your lawyer will endeavor to make some claim on your behalf that you testified today? Isn’t that true? “A. No.

“Q. It’s not. So, you’ve instructed your lawyer not to make any claims to that effect?

“A. No.”

On redirect examination, the following exchange ensued between Mambrino and Minto:

“Q. . . . [Defense] counsel asked you about testifying in this case, and you said that you had no expectation that anything would be given to you as a result of your testimony, right?

“A. Yes.

“Q. Was any specific plea agreement ever made out between you and the state for your testimony here?

“A. No. There was no agreement.

“Q. And [defense] counsel said, ‘oh, you’re just doing it to be a good citizen,’ right?

“A. Yes.

“Q. Why are you doing it? Why are you testifying in this case?

“A. Because it’s the right thing to do.”

The following exchange occurred on recross-exami- nation:

“Q. . . . [N]ow you’re talking because it’s the right thing to do, right?

“A. Yes.

***

“Q. . . . Again, I’m going to ask you, so we’re clear, because [Mambrino] said to you, ‘there’s no deals,’ she said to you on redirect.

“A. No deals.”

Finally, Minto’s testimony ended with this exchange: “Q. . . . [Y]ou’re telling this jury that you had no con- siderations for your testimony, right?

“A. Yes, I have considerations, but those are the details that will happen.”

Immediately following Minto’s testimony, Mambrino called Minto’s defense counsel, Cohen, to testify. On direct examination, the following exchange took place: “Q. . . . [D]id we work out any sort of written plea agree- ment between . . . Minto and the state . . . ?

“A. No.

“Q. Did we work out any other sort of agreement between the state . . . and . . . Minto?

“A. No. There was no agreement, verbal or written, regarding this case . . . .

“Q. Okay. . . . Did you approach the [prosecution] and ask if [it] wanted to use . . . Minto as a witness in this particular case?

“A. Yes. We had discussed that at sometime in the past. “Q. Okay. And, after we discussed that, did the [pros- ecution] tell you what [it] would do if . . . Minto came in and testified truthfully in this case?

“A. That there would be some consideration given to . . . Minto if he did testify truthfully in this case. But, beyond that, there was no agreement for any specific sentence [or] any reduction of the sentence.

“Q. So, when you say ‘consideration,’ what specifically do you mean by that?

“A. Well, I would hope that, when he’s standing before the judge, and he’s being sentenced, or we would work out some plea agreement that would take into consideration the fact that he did come here, he did testify truthfully, and that . . . he might get a break from the judge, in other words, get a lower sentence than he would have other- wise. That would be the hope, perhaps, the expectation, but that’s all it is.

“Q. But there hasn’t been anything specific worked out? “A. No. There never is in these situations.

“Q. That’s just the expectation, and that’s what the [prosecution] said [it] would indicate to the court [pre- siding over Minto’s criminal case] also at the time that he [pleaded] guilty?

“A. Yes. That’s correct.

“Q. Other than that, any other promises made to him? “A. No.”

After the petitioner’s defense counsel referenced Cohen’s testimony during closing argument, Mambrino argued in rebuttal: “None of [the prosecution witnesses has] an interest in the outcome of this case. There was no deal. There is no deal between the state and . . . Minto. You heard an officer of the court, his lawyer . . . Cohen, get up there and tell you there is no deal between the state and . . . Minto. The only one who has an interest in this case is [the petitioner], and he continually makes up stories to divert attention away from him[self] and [to] get everybody else involved in it.” After requesting and listening to the playback of Minto’s testimony, the jury found the petitioner guilty of murder and conspiracy to commit murder. See State v. Bryan, supra, 193 Conn. App. 287–88, 293–94. The petitioner received a total effective sentence of eighty years of incarceration. Id., 294. At Minto’s subsequent sentencing for the Manchester robbery, Mambrino informed the sentencing court that “the only ‘agreement,’ so to speak, if you want to call it [that], was that, at the time that . . . Minto, in fact, [pleaded] guilty after . . . the [criminal] trials [of the petitioner and Hall-Davis] were over, if he cooperated in those trials, and he testified truthfully in those trial[s], that I would, in fact, let the court know that. I would not be recommending any particular sentence in any way; I never did recommend any sentence in any way.” Mambrino also told the sentencing court that Minto “testified probably better than any witness I’ve ever seen testify, except for an expert witness.” Minto’s sentenc- ing court confirmed that the prosecution never made a recommendation for a particular sentence.

The petitioner later filed a petition for a writ of habeas corpus, which included Napue claims for Mambrino’s failure to correct Minto’s and Cohen’s false or substan- tially misleading testimony. At the habeas trial, which was held in 2023, Mambrino testified that, during the petitioner’s criminal case, she disclosed to the petition- er’s defense counsel that she would eventually inform the sentencing court presiding over Minto’s criminal case about Minto’s cooperation with the state and his testimony against the petitioner. Mambrino also testified that, in similar situations, when a cooperating witness had his or her own criminal charges pending, Mambrino would “offer” this incentive as “consideration,” and Minto in fact “received the consideration . . . .”

Nonetheless, Mambrino denied that this arrangement was a “deal,” such that she had an obligation under Napue to correct Minto’s testimony denying the existence of any deal. She conceded that there was a “promise” and that there had “always been an agreement” between the state and Minto, but she explained that a “specific deal” would have entailed “something written up, specifically what [Minto] would have testified to, what the exact promise was, what the outcome would be.” Mambrino testified that, in her agreement with Minto, “[t]here wasn’t anything specific with respect to a plea and an amount of [prison] time [Minto] would get,” so she did not consider it a “deal . . . .” Minto’s testimony that nothing had been “given” or “promised” to him, in Mambrino’s view, “wasn’t false testimony.” Similarly, Mambrino tes- tified, Minto’s statement that he was testifying against the petitioner because it was “the right thing to do” contained “nothing to correct.” Mambrino maintained that she did not correct Minto’s testimony because “[i]t all was fleshed out in the direct examination.”

Mambrino also defended her rebuttal closing argu- ment—in which she twice told the jury that there was no deal—by testifying: “There was no deal. There was an agreement. . . . There was no deal, as words [were] used.” Additionally, when the habeas court questioned her about the difference between a “deal,” a “promise,” and an “agreement,” Mambrino responded by instead discussing her conception of a “plea agreement.” When pressed to define the three specific words in the court’s question, Mambrino answered that she used the word “deal” at the petitioner’s criminal trial only “in a rebut- tal type of way” and the word “promise” in “a rebuttal type of a verbiage.”

The habeas court denied the petitioner’s petition for a writ of habeas corpus, holding in part that there was no Napue violation. Although the court determined that there was an agreement between the state and Minto, it nevertheless found that Minto’s testimony “about the scope of the agreement” was “truthful” and that “the jury was plainly made aware that Minto was told that something would be done in exchange for his testimony” against the petitioner. The habeas court further deter- mined that, even if Minto’s testimony about the nature of the agreement was false or substantially misleading, Mambrino corrected that testimony through the testi- mony of Cohen, who accurately explained the scope of Minto’s agreement with the state.

The petitioner, on the granting of certification, appealed to the Appellate Court from the habeas court’s judgment denying his petition for a writ of habeas corpus.

He then filed a motion to transfer his appeal to this court, which we granted.

The petitioner claims that Mambrino violated Napue by failing to correct the false or substantially mislead- ing testimony of Minto and Cohen concerning Minto’s agreement with the state. Specifically, he argues that Mambrino failed to correct Minto’s testimony that there were “[n]o deals” and Cohen’s testimony that there was “no agreement” between Minto and the state. The respon- dent, the Commissioner of Correction, argues that the testimony of Minto and Cohen was truthful and that, even if Minto’s testimony was false or substantially misleading, Mambrino corrected Minto’s testimony by questioning Cohen. “[D]ue process is . . . offended if the state, although not soliciting false evidence, allows it to go uncorrected when it appears. . . . If a government witness falsely denies having struck a bargain with the state, or substantially mischaracterizes the nature of the inducement, the state is obliged to correct the misconception. . . . Regardless of the lack of intent to lie on the part of the witness, Giglio [v. United States, 405 U.S. 150, 154, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972)] and Napue require the prosecutor to apprise the court when he or she knows that the witness is giving testimony that is substantially misleading.” (Internal quotation marks omitted.) Gomez v. Commis- sioner of Correction, 336 Conn. 168, 175, 243 A.3d 1163 (2020). “To establish a Napue/Giglio violation, then, the petitioner must demonstrate that the state’s witnesses provided material, false or substantially misleading testimony that the prosecutor failed to correct.” Id., 176. When determining whether a witness’ statement was false or substantially misleading, we consider the statement in the context of his entire testimony. See, e.g., Greene v. Commissioner of Correction, 330 Conn. 1, 16–22, 190 A.3d 851 (2018), cert. denied sub nom. Greene v. Semple, 586 U.S. 1167, 139 S. Ct. 1219, 203 L. Ed. 2d 238 (2019).

“Whether a prosecutor knowingly presented false or misleading testimony [in violation of a defendant’s due process rights] presents a mixed question of law and fact, with the habeas court’s factual findings subject to review for clear error and the legal conclusions that the court drew from those facts subject to de novo review.” (Internal quotation marks omitted.) Gomez v. Commis- sioner of Correction, supra, 336 Conn. 175; see also, e.g., United States v. Garcia, 793 F.3d 1194, 1207 (10th Cir. 2015) (“[w]e review for clear error the [D]istrict [C]ourt’s factual findings on the first two elements [of Napue],” namely, whether government witness committed perjury and whether prosecution knew that witness’ testimony was false), cert. denied, 577 U.S. 1088, 136 S. Ct. 860, 193 L. Ed. 2d 758 (2016). Specifically, the habeas court’s assessments of whether the testimony of a state witness was false or substantially misleading and, if so, whether a prosecutor knowingly failed to correct it are factual determinations subject to clear error review, whereas the court’s ultimate materiality assessment—whether there is a reasonable likelihood that the witness’ false or substantially misleading testimony could have affected the jury’s verdict of guilty—is a legal conclusion that we review de novo. See, e.g., Gomez v. Commissioner of Correction, supra, 176–78 and n.5; Gaskin v. Commis- sioner of Correction, 183 Conn. App. 496, 532 n.23, 193 A.3d 625 (2018); see also, e.g., United States v. Garcia, supra, 1207; United States v. Freeman, 650 F.3d 673, 678–80 (7th Cir. 2011); Drake v. Portuondo, 553 F.3d 230, 239, 241 (2d Cir. 2009); United States v. Madori, 419 F.3d 159, 169 (2d Cir. 2005), cert. denied, 546 U.S. 1115, 126 S. Ct. 1080, 163 L. Ed. 2d 898 (2006). We note that the record of the petitioner’s criminal trial was admitted as a full exhibit in the habeas court.

We first review the habeas court’s findings regarding the testimony of Minto and Cohen at the petitioner’s criminal trial. Initially, during direct examination, Mam- brino attempted to draw the truth out of Minto regard- ing his agreement with the state. Mambrino followed up twice on Minto’s denials of any “deals” by instead asking Minto whether a prosecutor had “told” him that he or she would inform Minto’s sentencing court of his cooperation with the state. Minto again responded that “[n]o deal was made.” Mambrino then asked Minto, “[n]o specific deal was made, but was that told to you?” Minto answered “[y]es” in response, thereby acknowledging that he had been told that his cooperation would be made known to the court presiding over his criminal trial at the time of his own sentencing, but offering nothing more.

This answer was substantially misleading, and Mam- brino left it uncorrected. Minto’s acknowledgment about being told that his cooperation with the state would be made known to the court presiding over his criminal trial at the time of his own sentencing—his least misleading answer on this topic—was not an acknowledgment by Minto of the existence of an agreement between him and the state. Minto “substantially mischaracterize[d] the nature of the inducement”; (internal quotation marks omitted) Gomez v. Commissioner of Correction, supra, 336 Conn. 175; by testifying that the benefit he received was not part of a “deal” or agreement, i.e., that the benefit was not an inducement at all. This testimony remained uncorrected even though Mambrino was “obliged to correct the misconception.” (Internal quotation marks omitted.) Gomez v. Commissioner of Correction, supra, 175. The fact that Mambrino attempted to draw truthful testimony out of Minto, though laudable, is of no legal moment.2 Cf., e.g., United States v. LaPage, 231 F.3d 488, 490–92 (9th Cir. 2000) (finding Napue violation even when prosecutor conceded during rebuttal argument that witness’ testimony was false). The habeas court’s finding that “the jury was plainly made aware that Minto was told that something would be done in exchange for his testimony” is unsupported by the record and is clearly erroneous. (Emphasis added.)

On cross-examination, Minto testified that there were “[n]o deals” and that “[n]othing had been promised to [him]”—false statements unto themselves. He further testified that Cohen was not going to tell Minto’s sen- tencing court about his cooperation with the state and ambiguously answered whether he instructed Cohen not to inform Minto’s sentencing court. Even if a hairsplit- ting distinction may be drawn here—based on whether Cohen, rather than Mambrino, was the specific person who would speak during Minto’s sentencing with the court presiding over his criminal case—Minto’s testi- mony was “in part false and otherwise so misleading as to amount to falsity”; United States v. Vozzella, 124 F.3d 389, 390 (2d Cir. 1997); because it cast doubt on the existence of any agreement in the presence of the jury. On redirect examination, Mambrino pivoted from trying to elicit correct testimony from Minto to instead shoring up his credibility. She asked him: “Was any specific plea agreement ever made out between you and the state for your testimony here?” (Emphasis added.) Minto not only denied the existence of a “specific plea agreement,” but he also testified that “[t]here was no agreement” whatsoever. Mambrino thus attempted to shore up Minto’s credibility by eliciting testimony about a different kind of agreement entirely, so that Minto could deny the existence of that agreement in front of the jury, and Minto doubled down by disavowing any and all agreements. Cf. Napue v. Illinois, supra, 360 U.S. 270–71.

On recross-examination, Minto again falsely testified that there were “[n]o deals” between him and the state. Then, when defense counsel asked Minto whether he was “telling this jury that [he] had no considerations for [his] testimony,” Minto answered, “[y]es, I have consider- ations, but those are the details that will happen.” This answer was Minto’s final testimony at the petitioner’s criminal trial. There is no surrounding context for what those “details” are, or what Minto meant by “consider- ations . . . .” Even if not false, this answer is substantially misleading because it immediately followed another denial of any “deals” and tends to ratify this previous falsehood. Furthermore, in Gomez v. Commissioner of Correction, supra, 336 Conn. 168, we cited Jenkins v. Artuz, 294 F.3d 284 (2d Cir. 2002), for its proposi- tion that a “tepid admission” by a witness during cross- examination was insufficient to cure the impact of false testimony on the jury, especially when the prosecutor sought to shore up the witness’ credibility. (Internal quotation marks omitted.) Gomez v. Commissioner of Correction, supra, 189, citing Jenkins v. Artuz, supra, 293–96.

On the basis of the totality of his testimony, the habeas court determined that Minto did not testify falsely, insofar as there was merely “some ambiguity in Minto’s testimony” due to the “different terms” that he and Mam- brino used, and that this ambiguity was later clarified. This finding is clearly erroneous due to Minto’s blanket denial of the existence of an agreement, which was never corrected. To the contrary, during redirect examina- tion, Mambrino reinforced the idea that Minto had no incentive to wrongly implicate the petitioner by having Minto confirm that he “had no expectation that anything would be given to [him] as a result of [his] testimony” and by eliciting from him both that he had “no agreement” with the state and that he was testifying because “it’s the right thing to do.” Given this testimony on redirect examination, any reliance by the habeas court on Minto’s final statement on recross-examination to support its conclusion that “the jury was plainly made aware” of the agreement was misplaced.3

The habeas court also relied on Cohen’s testimony as clarifying the nature of the agreement that Minto had with the state. For his part, Cohen testified that “[t]here was no agreement, verbal or written, regarding this case . . . .” For the reasons discussed previously, this unequivocal answer was, at a minimum, substantially misleading. Cohen admitted that “there would be some consideration given to . . . Minto if he did testify truth- fully” in the petitioner’s criminal case but insisted that “there was no agreement for any specific sentence [or] any reduction of the sentence.” As with Minto’s response to Mambrino’s question regarding “any specific plea agree- ment,” Cohen’s answer does not elucidate the nature of the actual agreement that existed in this case. For this reason, it is substantially misleading. Cohen further id., 13; despite the facts that the witness had entered a guilty plea prior to his testimony and that the transcript of the plea reflected that his cooperation with the state would be taken into consideration when he was sentenced. See id., 5–6, 9 and n.5, 12. This court agreed with the respondent that the witness’ “testimony [made] it clear that, when he testified that he had no ‘deal,’ he was not broadly denying that he had received any benefit in exchange for his testimony.” (Emphasis omitted.) Id., 14. After reviewing the questions asked of the witness on direct examination and cross-examination; see id., 16–20; we concluded: “It is evident that [the witness] was responding to questions regarding the length of the sentence he expected to receive, not whether he expected any benefit whatsoever. His response that he could ‘do the time’ fur- ther shows that [the witness] understood the question to be directed at the sentence he expected to receive. Because there was no agreement with respect to his specific sentence, [the witness’] testimony was not substantially misleading.” Id., 20–21. Central to our analysis was the fact that the witness’ “testimony made clear that he had received some benefit [from the state], namely, that he had pleaded to nonhomicide charges, which carry a significantly reduced sentence”; id., 22; thereby capping his exposure for his involvement in a shooting in which six people were shot, including one person who died as a result. See id., 5–6, 10, 17, 21–22.

Unlike the statements of the witness in Greene, Minto’s statements that there were “[n]o deals” and that there was “no agreement” between him and the state necessarily encompassed more than whether there was an agreement concerning what his specific sentence would be—despite Mambrino’s asking him whether they had entered into a “specific plea agreement . . . .” Moreover, Mambrino not only failed to clarify the existence of an agreement, but she also sought to capitalize on the absence of such an agreement on redirect examination of Minto and in her rebuttal argument, thereby reinforcing, rather than correcting, the misleading impression that Minto had received no inducement to testify. testified that there was no “consideration” for Minto, other than merely “the hope, perhaps, the expectation,” that the sentencing court presiding over Minto’s criminal case would “take into consideration” Minto’s cooperation with the state and testimony against the petitioner. This answer is also substantially misleading because it does not put the jury on notice that there had “always been an agreement” between Minto and the state, as Mambrino testified at the habeas trial, and as the habeas court deter- mined. Finally, the petitioner’s defense counsel asked Cohen regarding his “hope” or “expectation”: “That’s just the expectation, and that’s what the [prosecution] said [it] would indicate to the court [presiding over Minto’s criminal case] . . . at the time that he [pleaded] guilty?” Cohen answered affirmatively. This answer is substan- tially misleading because the jury remained unaware that it was much more than a “hope” or “expectation”—it was an agreement that, in Mambrino’s words, involved an “offer” of “consideration” to Minto. Indeed, accord- ing to Mambrino’s testimony, Minto “received the con- sideration” only after holding up his end of the deal by testifying for the state at the petitioner’s criminal trial. The habeas court clearly erred in its finding that Cohen had “accurately explained the extent of Minto’s agree- ment with the prosecution.”

The habeas court’s conclusion that Mambrino suffi- ciently clarified the nature of the state’s agreement with Minto through her questioning of Minto and Cohen is further belied by how Mambrino viewed the totality of the testimony and described it to the jury in her closing argument. She told the jury that Minto had no interest in the outcome of the case, despite being fully aware that Minto’s outstanding criminal charges would not be resolved until after the outcome of the petitioner’s criminal trial was known and the petitioner had been sen- tenced. Thus, Minto was aware when he testified against the petitioner that the prosecution would not make its statement to the court presiding over Minto’s criminal case until it knew the full value of his cooperation. More important, Mambrino did not argue to the jury that there was an agreement only to bring Minto’s cooperation with the state to the attention of his sentencing court, and no agreement with Minto as to a specific sentence. Instead, she told the jury: “There is no deal between the state and . . . Minto. You heard an officer of the court, his lawyer . . . Cohen, get up there and tell you there is no deal between the state and . . . Minto.”4 Mambrino, herself an officer of the court and a minister of justice, must have believed that the testimony of Minto and Cohen supported this argument, even though she knew that she had made a deal with Minto to bring Minto’s coop- eration with the state to the attention of his sentencing court in exchange for his testimony. In sum, given that Mambrino elicited substantially misleading testimony from Minto and Cohen, failed to correct the false impres- sions created by their misleading testimony, and relied on the misleading testimony during closing argument, we conclude that the habeas court’s findings that the testimony of Minto and Cohen was not substantially misleading and that, to the extent it was, Mambrino had corrected it, were clearly erroneous.5

These Napue violations are not cured simply because, as the habeas court noted, “[t]here is no claim that the [prosecution] failed to disclose the parameters of the agreement with Minto to the defense . . . .” As we held in Gomez v. Commissioner of Correction, supra, 336 Conn. 168, suppression in the Brady6 sense is not a requirement under Napue, and disclosure “does not necessarily discharge the prosecutor from his duty to correct the false testimony or immunize the state from a claim that the defendant’s right to due process was violated.” Id., 183. Instead, to determine whether there was a Napue violation despite the disclosure, we consider the following five nonexhaustive factors: “[1] whether it is the prosecution or the defense that elicits the false testimony, [2] whether and how the prosecutor adopts and uses the false testimony, [3] the importance of the witness and his or her false testimony to the state’s case, [4] whether—and to what effect—defense counsel tries to impeach the perfidious witness or whether counsel has a clear tactical reason for not doing so, and, most important, [5] whether the truth ultimately is revealed to the jury.” Id., 185. In this case, all five factors support the conclusion that there were Napue violations despite 336 Conn. 175. It appears that the habeas court assumed Mambrino’s knowledge, however, in its discussion of whether Mambrino corrected Minto’s testimony. It stated: “Even if Minto’s testimony was false or substantially misleading, and the [prosecution] had an obligation to correct that testimony, it did do so through the testimony of . . . Cohen . . . .” In its analysis, the court did not find that Mambrino’s knowl- edge—which triggers this Napue obligation—was a contested issue. The habeas court’s other factual findings support this interpretation. For example, it found that “Mambrino . . . made a promise . . . that, if Minto testified in accordance with the information he gave to the police and cooperated with the state during the [petitioner’s criminal] trial, his cooperation would be made known to the sentencing judge if he [pleaded] guilty in his own case.” In other words, Mambrino knew about the agree- ment because she executed it. This finding is consistent with Mambrino’s testimony at the habeas trial that Minto “received the consideration” that she “offered,” that there had “always been an agreement” between the state and Minto, and that she believed the agreement “was fleshed out in the direct examination.” The habeas court’s findings regarding Mambrino’s knowledge, coupled with its clearly erroneous finding that Mambrino corrected any false or substantially misleading testimony, provide the factual predicate to establish this element of Napue. any disclosure of the agreement, as the record shows that Mambrino elicited the false or substantially misleading testimony of Minto, Mambrino adopted and relied on that false or substantially misleading testimony both during Minto’s redirect examination and her closing argument, Minto and his testimony were important to the state’s case against the petitioner, and Mambrino’s use of the false or substantially misleading testimony frustrated defense counsel’s attempts at impeachment, keeping the truth from the jury.

What remains is to assess the materiality of these Napue violations. The petitioner argues that they are material because Minto’s testimony was an important, if not critical, part of the state’s case against the petitioner, and “[c]orrecting Minto’s false and misleading testimony would have undermined his credibility and demonstrated his willingness to lie under oath.” The respondent argues that “the state’s case was strong even without Minto’s testimony”7 and, therefore, that there is no reasonable likelihood that any false or misleading testimony could have affected the jury’s verdict.

The following rules govern our evaluation of a prosecu- tor’s failure to correct false or substantially misleading testimony: “[A prosecutor’s securing] a conviction with evidence that he or she knows or should know to be false . . . is fundamentally unfair, and [such a conviction] must be set aside if there is any reasonable likelihood that the false testimony could have affected the [verdict] of the jury. . . . This standard . . . is not substantively different from the test that permits the state to avoid having a conviction set aside, notwithstanding a violation of con- stitutional magnitude, upon a showing that the violation was harmless beyond a reasonable doubt. . . . This strict standard of materiality is appropriate in such cases not just because they involve prosecutorial [impropriety], but more importantly because they involve a corruption of the truth-seeking function of the trial process. . . . In light of this corrupting effect, and because the state’s use of false testimony is fundamentally unfair, prejudice sufficient to satisfy the materiality standard is readily shown . . . such that reversal is virtually automatic . . . unless the state’s case is so overwhelming that there is no reasonable likelihood that the false testimony could have affected the [verdict] of the jury.” (Citations omitted; footnotes omitted; internal quotation marks omitted.) Adams v. Commissioner of Correction, 309 Conn. 359, 371–73, 71 A.3d 512 (2013). “[I]t will be the unusual case in which the prosecutor fails to correct material, misleading testimony regarding the existence of a coop- eration agreement and a reviewing court can, neverthe- less, determine with confidence that the jury was not misled thereby.” Gomez v. Commissioner of Correction, supra, 336 Conn. 190.

The following facts are relevant to our analysis. Dur- ing the petitioner’s criminal trial, while Minto’s agree- ment with the state was in effect, “Minto’s testimony . . . directly implicated the [petitioner] in the victim’s murder.” State v. Bryan, supra, 193 Conn. App. 319. It was through Minto’s testimony that Hall-Davis’ hear- say statements concerning the victim’s murder were admitted at trial. See id., 294. These hearsay statements included Hall-Davis’ “confess[ion] to Minto that he had killed the victim at the [petitioner’s] behest in order to terminate the victim’s pregnancy.” Id., 291. Minto’s rendition of Hall-Davis’ confession thereby painted the petitioner as the mastermind of the murder. The Napue violations that hid the “inducement” of Minto’s agree- ment; State v. Ouellette, 295 Conn. 173, 186, 989 A.2d 1048 (2010); as a possible “motivation to lie”; (internal quotation marks omitted) id., 190; unfairly made Minto appear to be a more reliable conduit for the incriminating hearsay statements. See, e.g., Adams v. Commissioner of Correction, supra, 309 Conn. 372 (“the state’s use of false testimony is fundamentally unfair”).

Additionally, Minto’s testimony is the only evidence linking the murder weapon to the murder itself. See State v. Bryan, supra, 193 Conn. App. 292–93, 318.

Minto testified that the recovered .44 Ruger was the same gun that Hall-Davis had used in committing the Hartford murder and the Manchester robbery. See id. He also testified that Hall-Davis had disposed of a shell casing tying the firearm to the murder. See id., 292, 318. Furthermore, as the record of the petitioner’s criminal trial indicates, Minto provided the only account that the petitioner and Hall-Davis ultimately were able to fix the firearm after Lewis could not. Throughout his testimony, Minto continued to associate the murder weapon and Hall-Davis with the petitioner, stating that he “hear[d]” the petitioner become upset that Hall-Davis had not discarded the .44 Ruger.

Defense counsel made some progress in impeaching Minto with respect to the murder weapon, but the incon- sistencies in Minto’s testimony could have appeared less innocent if the jury knew about Minto’s agreement with the state. See, e.g., State v. Ouellette, supra, 295 Conn. 190 (“[o]nly through complete and candid disclo- sure of a witness’ interest can the jury accurately gauge the credibility of the testimony proffered”). On cross- examination, Minto admitted that he never personally saw the petitioner and Hall-Davis fix the .44 Ruger and that he never witnessed the gun successfully fire at any time, not even during the Manchester robbery. Minto acknowledged during cross-examination that the basis for his testimony that the .44 Ruger was fixed before the commission of the victim’s murder was that the petitioner and Hall-Davis were “happy” after trying to fix the gun themselves.

More significant, Minto did not admit on direct exami- nation that, when he “hear[d]” the petitioner’s heated conversation with Hall-Davis inside a car, Minto was only watching from outside of the car and could not in fact hear their conversation. Minto’s sole basis for what he “hear[d]” was Hall-Davis telling him about the conversation after it happened; it was more hearsay. On redirect examination, Minto clarified that he only saw the petitioner showing Hall-Davis a surveillance video of the Manchester robbery on his cell phone, which dis- played the .44 Ruger, and that the petitioner appeared to be upset. On recross-examination, however, Minto admitted that, for this to be possible, he would have had to view the petitioner’s cell phone screen through the car’s tinted windows, while Minto himself was sitting in a nearby car that also had tinted windows. If not for the Napue violations that enhanced Minto’s credibility dur- ing the petitioner’s criminal trial, the inconsistencies in Minto’s testimony could have come across as self-serving embellishments connecting the petitioner to the murder weapon rather than as good faith recollections. See, e.g., State v. Ouellette, supra, 295 Conn. 190–91.

We turn to the strength of the state’s remaining case against the petitioner, which is the respondent’s principal argument for why the Napue claims are not material. As the respondent puts it, “the state’s case was strong even without Minto’s testimony.” The respondent points to evidence that included Lewis’ testimony that the peti- tioner and Hall-Davis had approached him about fixing the .44 Ruger, and that, when Lewis indicated that he could not fix it, Hall-Davis told Lewis that he would repair it; the testimony of the victim’s father that, on the evening before the early morning murder, the petitioner and Hall-Davis had attended a cookout at his home and left together; Walker’s testimony that, less than two hours before the victim’s murder, the petitioner, with Hall-Davis accompanying him, had told Walker that he wanted to kill the victim; Walker’s testimony that the petitioner had asked him to act as a lookout and to pro- vide a false statement to the dispatched police officers; Walker’s testimony that the petitioner had told him that he wanted to kill the victim because she would not get an abortion; and the petitioner’s statement to the police, indicating that he knew that the victim was pregnant, that he was angry about it, and that he had asked her numerous times to get an abortion. The respondent also cites several inconsistencies in the petitioner’s state- ments, including the petitioner’s statement that Hall- Davis entered the backseat of the victim’s car, even though the car had only two doors.

Walker also testified, however, that the police were “threatening” his freedom and that he “felt” like the police were “going to put [him] in jail” if he did not share information regarding the circumstances surround- ing the victim’s murder. He stated under oath that his landlord had told him that Detective Early came to his apartment looking for him and that he had no choice but to be driven directly to the police station because “the police [were] going to kick down [his] door if [he] didn’t come down there.” Walker further testified that, when he was at the police station, he and Detective Early were “debating” the contents of his sworn statement and, perhaps most important, that he acceded to Detective Early’s version of events that he himself had witnessed. Walker stated that he heard “one or two” gunshots on the night of the shooting, “but [Detective Early] put two” gunshots in Walker’s statement, so he told Detective Early, “all right, two.”

After accounting for the effect of the Napue viola- tions, we are not persuaded that the remainder of the state’s case against the petitioner was “overwhelming . . . .” (Internal quotation marks omitted.) Adams v. Commissioner of Correction, supra, 309 Conn. 372–73. The remaining evidence that the respondent marshals is that the petitioner attempted to fix a .44 Ruger with Hall-Davis before the victim’s murder, the petitioner was at the home of the victim’s father with Hall-Davis just hours before the murder, the petitioner was frustrated that the victim would not seek an abortion, the petitioner gave inconsistent statements to the police, and Walker’s testimony established the petitioner’s motive and plan for the murder and his conspiracy to commit the murder. Under Napue’s materiality standard, we cannot con- clude that the jury would have found Walker’s testi- mony sufficient to carry the weight of this conviction for murder and conspiracy to commit murder and to render the Napue violations “harmless beyond a reasonable doubt.” Id., 372. We therefore conclude that there is a “reasonable likelihood that the false [or substantially misleading] testimony could have affected” the jury’s guilty verdict. (Internal quotation marks omitted.) Id., 371.8

Mambrino’s argument before the jury worsened this prejudice. See, e.g., Glossip v. Oklahoma, 604 U.S. 226, 250, 145 S. Ct. 612, 221 L. Ed. 2d 90 (2025) (“[a]lthough the prosecution’s failure to correct [the witness’] false testimony was a material Napue violation on its own, additional conduct by the prosecution further under- mines confidence in the verdict”). Mambrino told the jury: “None of [the prosecution witnesses has] an interest in the outcome of this case. There was no deal. There is no deal between the state and . . . Minto. You heard an officer of the court, his lawyer . . . Cohen, get up there and tell you there is no deal between the state and . . . Minto. The only one who has an interest in this case is [the petitioner], and he continually makes up stories to divert attention away from him[self] and [to] get everybody else involved in it.” As in Gaskin v. Commissioner of Correc- tion, supra, 183 Conn. App. 496, “the prosecutor . . . bolstered [the witness’] credibility by suggesting to the jury that [the witness] had nothing to gain, sharpening the already substantial prejudice.” Id., 524 n.17; see also, e.g., Gomez v. Commissioner of Correction, supra, 336 Conn. 191 (“in his closing argument, not only did [the prosecutor] not correct the witnesses’ misstatements, but he affirmatively vouched for [a witness’] credibility and invited the jury to decide the case on the basis thereof”). Mambrino later explained to the habeas court that she used the word “deal” during the petitioner’s criminal trial only “in a rebuttal type of way” and the word “promise” in “a rebuttal type of a verbiage.” Mambrino’s use of the Napue violations during her closing argument reflects an approach to the truth that further undermines con- fidence in the verdict.

The judgment of the habeas court is reversed and the case is remanded to that court with direction to grant the petitioner’s third amended petition for a writ of habeas corpus, to vacate the petitioner’s conviction of murder and conspiracy to commit murder, and to order a new trial.

In this opinion MULLINS, C. J., and D’AURIA, ECKER and BRIGHT, Js., concurred.

1 For this reason, we have no occasion to reach the petitioner’s claims brought under Brady v. Maryland, 373 U.S. 83, 87, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963). Notwithstanding that we need to go no further than our Napue analysis, we would be remiss if we did not acknowledge the concurring and dissenting opinion’s thoughtful and thorough treat- ment of the Brady claims. We see no flaws in its analysis, and, had we not disposed of this case under Napue, we would have been inclined to embrace its approach. 2 Mambrino had other options available to her when Minto failed to acknowledge that he had a cooperation agreement with the state. For example, she could have asked the trial court to instruct the jury that the parties stipulate to the existence of such an agreement. See, e.g., United States v. LaPage, 231 F.3d 488, 492 (9th Cir. 2000) (“[m]any prosecutors, when this occurs, interrupt their own questioning, and work out in a bench conference with the judge and defense counsel how to inform the jury immediately that the testimony is false”). 3 The facts of this case are distinguishable from those in Greene v. Commissioner of Correction, supra, 330 Conn. 1. In Greene, this court determined that a cooperating witness’ testimony that he had “no deal” with the state and “was expecting nothing in return for his testimony” was not substantially misleading; (internal quotation marks omitted) 4 During her testimony at the petitioner’s habeas trial, Mambrino seemed to justify her statement in closing argument that there was no “deal” by distinguishing an “agreement,” which she testified she had with Minto, from a “deal.” For example, she testified: “There was no deal. There was an agreement. . . . There was no deal, as words [were] used.” According to Mambrino, there is a “deal” only when there is a specific agreement as to charges and a sentence. We find such a justifica- tion perplexing. In ordinary parlance, and as a matter of contract law, a “deal” and an “agreement” are one and the same. An “agreement” is defined in relevant part as “an arrangement as to a course of action”; Merriam-Webster’s Collegiate Dictionary (12th Ed. 2026) p. 37; and a “deal” is defined as “an arrangement for mutual advantage . . . .” Id., p. 423. To the extent that Mambrino believed there was a difference between the two, she never communicated that to the jury. Thus, when she said in her closing argument that there was no “deal” with Minto, the jury would have understood that to mean no “agreement,” which clearly was not true. 5 Given the habeas court’s findings, it did not expressly assess whether Mambrino knowingly failed to correct the false or substantially mislead- ing testimony. See, e.g., Gomez v. Commissioner of Correction, supra, 6 Brady v. Maryland, 373 U.S. 83, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963). 7 The respondent also argues that the petitioner’s defense counsel successfully impeached Minto despite any Napue violations. For the reasons previously discussed, we disagree. 8 We observe that the Appellate Court upheld the petitioner’s convic- tion on direct appeal based in part on Minto’s testimony. Faced with a claim under Brady v. Maryland, 373 U.S. 83, 87, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963), concerning the suppression of impeachment evidence relating to Detective Early; see State v. Bryan, supra, 193 Conn. App. 288, 306; the Appellate Court was “not convinced that impeaching [his] credibility . . . would have impacted the jury’s consideration of Minto’s testimony, which directly implicated the [petitioner] in the victim’s murder.” Id., 319. That Minto’s testimony already has undergirded an affirmance of the petitioner’s conviction provides further support for our conclusion that the Napue violations relating to Minto were material.

Case-law data current through September 30, 2026. Source: CourtListener bulk data.