18-1392-cr United States v. Wedd

In the United States Court of Appeals for the Second Circuit

August Term, 2020 No. 18-1392-cr

UNITED STATES OF AMERICA, Appellee,

v.

DARCY WEDD, Defendant-Appellant. ∗

Appeal from the United States District Court for the Southern District of New York. No. 1:15-cr-616 — Katherine B. Forrest, Judge.

ARGUED: DECEMBER 10, 2020 DECIDED: APRIL 1, 2021

Before: CABRANES, PARK, and NARDINI, Circuit Judges.

∗ The Clerk of Court is directed to amend the caption as set forth above. Defendant-Appellant Darcy Wedd appeals from a judgment of conviction entered on May 3, 2018, following a jury trial in the United States District Court for the Southern District of New York (Katherine B. Forrest, J.). Wedd argues that the district court erred by (1) failing to recuse itself under

(a); (2) giving a jury instruction on conscious avoidance; and (3) allowing the Government to inadequately plead and prove identity theft under 18 U.S.C. § 1028A. We hold that the district court’s conduct did not create an appearance of partiality warranting recusal under Section 455(a). As to the charge, the district court properly instructed the jury on conscious avoidance because sufficient trial evidence supported that theory of criminal liability. Finally, we conclude that Wedd’s conduct, as pled and as proven at trial, fit squarely within the scope of the aggravated identity theft statute. We therefore AFFIRM the decision of the district court.

MARC FERNICH, New York, New York, for Defendant-Appellant. RICHARD COOPER (Anna M. Skotko, on the brief), Assistant United States Attorneys, for Audrey Strauss, Acting United States Attorney for the Southern District of New York, New York, New York, for Appellee.

WILLIAM J. NARDINI, Circuit Judge:

This case involves a technological fraud. Defendant-Appellant Darcy

Wedd and his co-conspirators used a computer program to automatically

subscribe cell phone users, without their knowledge or consent, to paid text

2 message services. After two trials resulted in deadlocked juries, the third

trial resulted in a conviction for Wedd.

Wedd now appeals from a judgment of conviction entered May 3,

2018, in the United States District Court for the Southern District of New

York (Katherine B. Forrest, J.). He argues that the district court erred by:

(1) failing to recuse itself under

(a) before his third trial;

(2) giving a jury instruction on conscious avoidance; and (3) allowing the

Government to inadequately plead and prove aggravated identity theft

under 18 U.S.C. § 1028A.

We first hold that the district court’s conduct did not create an

appearance of partiality warranting recusal under Section 455(a). Next, we

conclude that the district court properly charged the jury on conscious

avoidance because sufficient trial evidence supported that theory of

criminal liability. We further conclude that Wedd’s indictment adequately

pled violations of Section 1028A and so the district court properly declined

to dismiss the Section 1028A counts. Finally, we conclude that the district

3 court properly denied Wedd’s motion for acquittal because a rational jury

could have concluded beyond a reasonable doubt that his conduct, as

proven at trial, fit squarely within the scope of the aggravated identity theft

statute: By participating in a scheme that employed victims’ cell phone

numbers to sign them up for paid text message services without their

knowledge or consent by means of an auto-subscribing computer program,

Wedd “use[d], without lawful authority, a means of identification of

another person” within the meaning of Section 1028A. We therefore

AFFIRM the judgment of the district court.

I. Background

A. The offense conduct

On June 5, 2017, the Government filed an eight-count superseding

indictment (the “Indictment”) charging Wedd with various crimes

stemming from his role in two schemes to automatically subscribe

4 consumers, without their knowledge or consent, to premium SMS text

messaging services (“PSMS Services”). 1

PSMS Services provide subscribers with recurring cell phone text

messages containing content such as celebrity gossip, IQ quizzes, stock tips,

and daily horoscopes. Subscription results in a regular monthly charge

(typically $9.99) on consumers’ phone bills.

A double opt-in, or two-factor, verification process is designed to

prevent fraud and ensure that only willing consumers pay for subscriptions

to PSMS Services. Consumers typically initiate the subscription process by

entering their telephone number into a website. The consumers then receive

a text message with a verification code, which the consumers enter back into

the website (or confirm by replying to the text message). The consumers

1“SMS” stands for “short message service” and “is what is commonly known as a text message; although SMS messaging only allows users to send and receive messages of up to 160 alpha-numeric characters.” United States v. Streett,

, 1253 n.28 (D.N.M. 2018) (internal quotation marks and alteration omitted). “SMS” is distinct from “MMS,” which stands for “multimedia messaging service” and “includes pictures, forty seconds or less of video, audio, and text messages greater than 160 alpha-numeric characters in length.”

at 1245 n.21.

5 then get a welcome text message and start getting charged for the services.

Both of the schemes at issue here—as described in further detail below—

involved manipulating this verification process to enroll unsuspecting

customers in PSMS Services.

During the period covered by the Indictment, Wedd served as the

Chief Operating Officer and then Chief Executive Officer of a company

called Mobile Messenger. Mobile Messenger was a mobile aggregator—that

is, an intermediary between the digital content providers that market PSMS

Services to consumers and the mobile phone carriers that transmit the

messages to consumers. Mobile aggregators compile consumers’ monthly

PSMS Services charges and bundle them for inclusion on the phone bills sent

by carriers to consumers, receiving approximately 25% of the revenue

generated.

The Indictment alleged that, between 2011 and 2013, Wedd worked

with two sets of content providers—Tatto Media, Inc. (“Tatto”) and certain

companies operated by Wedd’s co-defendant Eugeni Tsvetnenko, also

6 known as “Zhenya”—to bypass 2 the double opt-in verification procedures

using an “auto-subscribing” computer process that made it look as if

consumers had provided consent to PSMS Services. Monthly charges

would then appear on consumers’ bills until the consumers noticed them

and tried to cancel. Cancellation was often difficult.

The first four counts of the Indictment related to the Tatto scheme; the

next four related to the Tsvetnenko scheme. Counts One and Five of the

2 Wedd describes this activity as “spoofing.” See, e.g., Def. Br. at 4, 41-42. That term dates back to the late 1800s, when an English actor and comedian named Arthur Roberts devised a game entitled “Spoof.” See Definition of Spoof, MERRIAM WEBSTER, https://www.merriam-webster.com/dictionary/use. The term acquired a meaning akin to carrying out a hoax—or, depending on context, engaging in parody. See Spoof: It’s All Fun and Games Until Somebody Gets Spoofed, MERRIAM WEBSTER, https://www.merriam- webster.com/words-at-play/spoof-meaning-origin. The word has acquired particular currency in recent years in connection with e-mail, Medidata Sols., Inc. v. Fed. Ins. Co.,

(S.D.N.Y. 2017) (defining e-mail “spoofing” as “the practice of disguising a commercial e-mail to make the e-mail appear to come from an address from which it actually did not originate” (quotation marks omitted)), aff’d,

(2d Cir. 2018), and telephone numbers, United States v. Sayer,

(1st Cir. 2019) (describing “spoofing” applications on a defendant’s phone that “enabled him to place outgoing phone calls under the guise of a different phone number”). The term has even acquired a special meaning in commodities trading fraud. 7 U.S.C. § 6c(a)(5)(C) (defining “spoofing” under the Commodities Exchange Act as “bidding or offering with the intent to cancel the bid or offer before execution”). In the interest of precision, we will avoid use of the term “spoofing” in this opinion.

7 Indictment charged Wedd with conspiracy to commit wire fraud in

violation of

. Counts Two and Six charged Wedd with wire

fraud in violation of

and 2. Counts Three and Seven

charged Wedd with aggravated identity theft in violation of 18 U.S.C.

§§ 1028A and 2. Counts Four and Eight charged Wedd with conspiracy to

commit money laundering in violation of

(h).

B. Mistrial after mistrial

On April 3, 2017, a trial began for Wedd and several co-defendants.

The jury deadlocked on all the defendants, so the district court declared a

mistrial on May 3, 2017.

On August 15, 2017, a second trial began for Wedd and one co-

defendant, Fraser Thompson. On September 5, 2017, the jury convicted

Thompson on all counts but again was deadlocked as to the charges against

Wedd. Accordingly, the district court declared a mistrial as to Wedd for the

second time.

8 After the second trial, the district court talked to the jury and reported

to the parties that the vote had been 11-1 in favor of conviction. The district

court later received a letter from a juror in the majority, expressing

frustration with the holdout juror.

C. Wedd’s reassignment request

On the same day that the district court declared the second mistrial,

the court—unprompted—queried whether “either side” might “believe[]

that [it] would benefit from a different judge” for the third trial. App’x at

87. The district court raised the question of presiding over the third trial

itself. See

(“I would ask that if at this point, given the fact that the Court

has tried this case now twice, I would try the case a third time.”). The court

speculated that “the government may not have agreed with many of my

rulings on allowing in certain things, nor might the defendant” and that

either party “may want . . . [an] opportunity to raise things in front of

another judge.”

In the event of a reassignment request, the district court

stated that it would “consult with [its] colleagues” and that “it may be that

9 it would be the right thing to do . . . to put [the case] in the wheel” for

reassignment.

After setting a date for Wedd’s retrial, the district court excused him

and his counsel. The district court proceeded to discuss the upcoming

sentencing of Wedd’s now-convicted co-defendant Thompson and his

counsel, previewing “some of the things that are important to [the court] at

the time of sentencing.”

. The district court identified the potential

for recidivism as its “biggest issue” in “sentencing a defendant in a fraud

case,” observing that “[t]here is an extremely high rate of recidivism with

fraud.”

The court noted, however, that “of all of the defendants,” it

believed Thompson “was least likely to” recidivate.

The court said: “[I]f

I were to rank people in terms of their levels of culpability, he would be at

the very bottom of all of these defendants, . . . and far below Mr. Wedd.”

.

On September 13, 2017, Wedd’s counsel filed a letter asking for a new

judge to preside over the third trial. The letter argued that there was

10 “substantial support for the idea of reassigning a case in the context of a

retrial following appellate reversal, where many of the same considerations

are in play” and that there was “no reason to believe” reassignment would

waste or duplicate judicial resources. No. 15 Cr. 616, Dkt. 497 at 2-3. The

letter added that, if the case were reassigned, “this Court will be free to make

necessary factual determinations in connection with Mr. Thompson’s

sentence, without concern that they may reflect upon the Court’s view of

Mr. Wedd’s alleged involvement.”

(footnote omitted). The

Government opposed the reassignment request “for purposes of judicial

efficiency.” No. 15 Cr. 616, Dkt. 500.

On September 28, 2017, the district court denied the reassignment

request, citing judicial efficiency. At a later pretrial conference, the district

court explained that it had spoken to other judges about the possibility of

reassignment, but none of them had time available in the period previously

discussed for the trial.

11 D. Wedd’s third trial

On December 4, 2017, Wedd’s third trial began. Key testimony for the

Government’s case came from three cooperating witnesses: Tatto co-owner

Lin Miao and Mobile Messenger executives Michael Pajaczkowski and

Erdolo Eromo. Wedd testified in his own defense. The cooperating

witnesses described Wedd as a knowing and active participant in the Tatto

and Tsvetnenko fraud schemes, while Wedd claimed he lacked knowledge

of either scheme.

1. Evidence of Wedd’s knowing participation in the Tatto scheme

Miao testified that, in 2011, Wedd informed Miao that an internal

Mobile Messenger audit had detected Tatto’s involvement in auto-

subscribing. By email, Wedd told Andrew Bachman, Tatto’s other co-

owner, that Tatto would be suspended from signing up new customers.

Wedd did not, however, require Tatto to stop collecting money from

already-billed subscribers. Nor did Wedd alert law enforcement, phone

carriers, or victim consumers about the detection of fraud.

12 Miao testified that he decided to address the matter with Wedd

“directly.” Tr. at 755. 3 Miao had “a very open conversation” with Wedd in

which Miao admitted Tatto had been auto-subscribing and Wedd “wasn’t

surprised.” Tr. at 759. Miao asked Wedd if Mobile Messenger could help

Tatto with auto-subscribing in exchange for a cut of the scheme’s proceeds.

Wedd agreed, saying “it would be a win for all parties,” and Miao and

Wedd went on to discuss how much money they could make and how to

conceal the fraud. Wedd instructed Miao to speak with Pajaczkowski at

Mobile Messenger about the scheme and, in particular, about how to hide

it.

Pajaczkowski testified that he and Miao then made a deal in which

they would work together to facilitate the auto-subscribing, with a portion

of the proceeds going to Wedd. Pajaczkowski and Wedd talked about

Wedd’s cut, and Wedd said he would like ten percent. Wedd also said that

3 “Tr.” refers to the transcript from Wedd’s third trial.

13 he had separately arranged for Tatto to give him a Rolex watch in exchange

for Mobile Messenger’s auto-subscribing assistance.

Pajaczkowski described how he and another Mobile Messenger

employee gave Tatto lists of phone numbers to be auto-subscribed, as well

as technical advice about how to hide the fraud from phone carriers and

regulators. Pajaczkowski also gave Wedd his share of the proceeds. (The

Government offered detailed financial records showing these payments.)

Pajaczkowski further testified that he and Wedd discussed how much

money they were making through the scheme, as well as “the need to be

very careful with Tatto.” Tatto employees acted recklessly at times. Once,

Tatto auto-subscribed a manager at Verizon, which Pajaczkowski testified

was “the equivalent of selling drugs to [an] undercover cop” and “about as

stupid and as bad as it gets.” Tr. at 295. Mobile Messenger’s relationship

with Tatto continued but, in February 2012, Pajaczkowski stopped giving

Tatto new cellphone numbers (but let Tatto continue defrauding already

auto-subscribed consumers).

14 2. Evidence of Wedd’s knowing participation in the Tsvetnenko scheme

Eromo testified that, in 2012, Mobile Messenger decided to enter into

a similar auto-subscribing scheme with certain of Tsvetnenko’s content

provider companies, CF Enterprises Pty Ltd and DigiMobi Pty Ltd. Eromo

described a meeting he had with Pajaczkowski and Wedd in which they

discussed possible auto-subscribing partners and agreed to approach

Tsvetnenko.

Pajaczkowski testified that Wedd and Tsvetnenko were “close

personal friends,” and that Wedd knew that Tsvetnenko “had been in

trouble a number of times for conducting auto-subscription schemes in

different countries.” Tr. at 371. Pajaczkowski further testified that he,

Wedd, Eromo, and Thompson discussed how to coordinate the auto-

subscribing scheme with Tsvetnenko and divided up necessary tasks

amongst themselves. Wedd assumed an “oversight” role, and the men

agreed to an even four-way split of the proceeds from the scheme. Tr. at

398-99.

15 Pajaczkowski testified that he obtained phone numbers from Mobile

Messenger’s database to give to Tsvetnenko for auto-subscribing purposes.

In return, Tsvetnenko wired proceeds from the scheme back to

Pajaczkowski and Eromo through shell companies. Pajaczkowski then paid

half of what he received to Wedd, and Eromo paid half of what he received

to Thompson. This scheme continued through the end of 2013.

3. Wedd’s testimony denying his knowing participation in the schemes

Wedd testified in his own defense, claiming that Pajaczkowski and

Eromo had actually agreed to participate in the fraud without Wedd’s

involvement. Wedd acknowledged that Miao and his business partner had

suggested that Mobile Messenger engage in auto-subscribing with Tatto.

However, Wedd testified that he rejected this idea, telling Miao and his

partner that this conduct was fraudulent and illegal. When Miao and his

partner asked to speak with Pajaczkowksi, Wedd claimed to have

responded, “You can talk to whoever the f[*]ck you want. You’re not auto-

subscribing on my platform.” Tr. at 1599-1600.

16 Wedd further testified that his frequent travels for work kept him

from noticing the fraud schemes. He also claimed that he would never have

agreed to participate in the fraud because it would have endangered his

legitimate business efforts.

4. Wedd’s conviction and sentence

On December 15, 2017, Wedd’s third trial ended with a conviction on

all counts.

On April 2, 2018, the district court sentenced Wedd to concurrent

terms of 72 months in prison on the fraud, conspiracy, and money

laundering counts, as well as mandatory consecutive terms of 24 months in

prison on each of the aggravated identity theft charges. Judgment entered

on May 3, 2018. This appeal followed.

II. Discussion

On appeal, Wedd argues that the district court erred by failing to

recuse itself under Section 455(a) despite his reassignment request; giving a

jury instruction on conscious avoidance; and allowing the Government to

17 inadequately plead and prove aggravated identity theft under Section

1028A. We address each of these arguments in turn.

A. The district court did not err, much less plainly err, in declining to recuse itself.

Wedd first argues that the district court improperly failed to recuse

itself in violation of

(a). Section 455(a) provides that “[a]ny

justice, judge or magistrate of the United States shall disqualify himself in

any proceeding in which his impartiality might reasonably be questioned.”

We review a district court’s decision not to recuse itself for abuse of

discretion. See LoCascio v. United States,

(2d Cir. 2007). We

will “rare[ly]” disturb a district court’s decision not to recuse itself. ISC

Holding AG v. Nobel Biocare Fin. AG,

(2d Cir. 2012) (internal

quotation marks omitted).

We evaluate partiality under Section 455(a) “on an objective basis, so

that what matters is not the reality of bias or prejudice but its appearance.”

Liteky v. United States,

(1994); see also Liljeberg v. Health Servs.

Acquisition Corp.,

(1988) (“The goal of section 455(a) is to

18 avoid even the appearance of partiality.” (internal quotation marks

omitted)). 4 We consider “whether a reasonable person, knowing all the

facts, would conclude that the trial judge’s impartiality could reasonably be

questioned.” United States v. Thompson,

(2d Cir. 1996)

(internal quotation marks and alteration omitted). “However, to say that

§ 455(a) requires concern for appearances is not to say that it requires

concern for mirages.” United States v. El-Gabrowny,

(S.D.N.Y. 1994) (Mukasey, J.). In close cases, “the balance tips in favor of

recusal.” Ligon v. City of New York,

(2d Cir. 2013) (internal

quotation marks omitted), vacated in part on other grounds,

(2d

Cir. 2014).

For purposes of Section 455(a), “‘[p]artiality’ does not refer to all

favoritism, but only to such as is, for some reason, wrongful or

4In 1974, Congress modified Section 455 to introduce an objective standard. “[T]he pre- 1974 law required a judge to recuse himself only when it was ‘improper, in his opinion, for him to sit.’” United States v. Bayless,

(2d Cir. 2000) (emphasis added) (quoting

(1970)). Under the post-1974 law, “the judge’s own subjective perception of impropriety is not necessary to invoke the statute.”

19 inappropriate.” Liteky,

. Accordingly, “opinions formed by

the judge on the basis of facts introduced or events occurring in the course

of the current proceedings, or of prior proceedings, do not constitute a basis

for a . . . partiality motion unless they display a deep-seated favoritism or

antagonism that would make fair judgment impossible.”

; see also

In re J.P. Linahan, Inc.,

(2d Cir. 1943) (“Impartiality is not

gullibility. Disinterestedness does not mean child-like innocence. If the

judge did not form judgments of the actors in those court-house dramas

called trials, [the judge] could never render decisions.”). 5

Ordinarily, Section 455(a) will not require recusal based on “a judge’s

comments during a proceeding that are critical or disapproving of, or even

5Cf. Liteky,

(“The judge who presides at a trial may, upon completion of the evidence, be exceedingly ill disposed towards the defendant, who has been shown to be a thoroughly reprehensible person. But the judge is not thereby recusable for bias or prejudice, since [the judge’s] knowledge and the opinion it produced were properly and necessarily acquired in the course of the proceedings, and are indeed sometimes (as in a bench trial) necessary to completion of the judge’s task. . . . It has long been regarded as normal and proper for a judge to sit in the same case upon its remand, and to sit in successive trials involving the same defendant.”).

20 hostile to, counsel, the parties, or their cases.” United States v. Carlton,

(2d Cir. 2008) (internal quotation marks omitted). Partiality

cannot be established through “expressions of impatience, dissatisfaction,

annoyance, and even anger, that are within the bounds of what imperfect

men and women, even after having been confirmed as federal judges,

sometimes display.” Liteky,

. Moreover, “[a] judge’s

ordinary efforts at courtroom administration—even a stern and short-

tempered judge’s ordinary efforts at courtroom administration—remain

immune” from challenge under Section 455(a).

. Thus, “the

standard for establishing a federal judge’s partiality, based on comments

made at trial, is quite difficult for a criminal defendant to meet.” Francolino

v. Kuhlman,

(2d Cir. 2004).

Wedd raises a Section 455(a) argument for the first time in this appeal.

He did seek reassignment below and, in doing so, stated that reassignment

would leave the district court “free to make necessary factual

determinations in connection with Mr. Thompson’s sentence, without

21 concern that they may reflect upon the Court’s view of Mr. Wedd’s alleged

involvement.” 15 Cr. 616, Dkt. 497 at 2-3 (footnote omitted). But Wedd

never argued that the district court, through any of its previous statements

regarding Thompson’s sentencing or any of its other conduct in connection

with the case, had displayed an inability to consider the case impartially.

Nor did Wedd invoke Section 455(a) at all below, or frame his request for

reassignment in any way around an impropriety in the district court

continuing to preside over the case. Accordingly, we review Wedd’s Section

455(a) challenge for plain error. See Carlton,

(“When [a

recusal] motion was not made below or a new ground for recusal is raised

on appeal, we review a district court’s failure to recuse itself for plain

error.”).

Federal Rule of Criminal Procedure 52(b) “permits an appellate court

to recognize a plain error that affects substantial rights, even if the claim of

error was not brought to the district court’s attention.” United States v.

Boyland,

(2d Cir. 2017) (internal quotation marks omitted).

22 Applying plain error review, the Court may “correct an error not raised at

trial only where the appellant demonstrates” the following:

(1) there is an error; (2) the error is clear or obvious, rather than subject to reasonable dispute; (3) the error affected the appellant’s substantial rights, which in the ordinary case means it affected the outcome of the district court proceedings; and (4) the error seriously affects the fairness, integrity or public reputation of judicial proceedings.

(internal quotation marks and alterations omitted).

We have held that “a judge’s failure to recuse himself might in some

circumstances” constitute plain error. Bayless,

. Here,

however, Wedd fails to satisfy the first prong of the plain error analysis—

the existence of an error. He claims that five aspects of the district court’s

conduct created an appearance of partiality warranting recusal under

Section 455(a). As to all five, whether singly or in combination, we disagree.

First, Wedd stresses that the district court sua sponte raised the

possibility of reassignment. However, nowhere in the district court’s

discussion of reassignment did it display any partiality towards the

Government. Nor can the court’s brief comments be read to suggest that the

23 court itself feared that it harbored feelings of partiality. The district court

simply suggested general reasons why either party might, in the ordinary

course of a mistrial, have a preference for a new judge. The district court

certainly did not suggest that the parties were entitled to one.

Second, Wedd argues that the district court’s brief statement

comparing the culpability of Thompson and Wedd created the appearance

of partiality. But Wedd does not argue, nor does the record suggest, that

the district court formed an opinion of Wedd’s culpability based on

anything other than “facts introduced or events occurring in the course of

the current proceedings, or of prior proceedings.” Liteky,

.

And the district court’s isolated reference to Wedd’s culpability does not

demonstrate “a deep-seated favoritism or antagonism that would make fair

judgment impossible.” Id.; cf. Carlton,

(even after “[h]aving

heard evidence and made a determination of [a] defendant’s guilt in a

revocation hearing, a judge may properly preside over the subsequent

criminal trial for the same offense”); McMahon v. Hodges,

, 290

24 (2d Cir. 2004) (even though “the trial judge had undoubtedly formed

opinions about [the defendant’s] likely guilt during the course of [a co-

conspirator’s] trial at which the judge presided,” recusal was not required).

Judges go to great lengths to maintain decorum in the courtroom and,

to the extent possible, to avoid signaling any particular view of an

unconvicted criminal defendant’s guilt. But judges are frequently, and quite

properly, required to make assessments of a defendant’s culpability before

a jury has returned a verdict. For example, judges often must consider the

strength of the evidence against a defendant when considering questions of

detention or release. See

(g)(2) (“The judicial officer shall, in

determining whether there are conditions of release that will reasonably

assure the appearance of the person as required and the safety of any other

person and the community, take into account the available information

concerning . . . the weight of the evidence against the person . . . .”); see also

United States v. Orena,

(2d Cir. 1993) (reversing district

court’s decision to release defendant from pretrial detention where strong

25 evidence ”demonstrated” defendant’s key role in violent criminal

organization).

Assessments of culpability can also become necessary mid-trial. We

have instructed district judges that before admitting statements of a co-

conspirator, they must make certain factual findings by a preponderance of

the evidence that may indicate the defendant’s guilt of the charged offense.

United States v. Geaney,

(2d Cir. 1969); see United States v.

Tracy,

(2d Cir. 1993) (“[S]tatements that would otherwise

be hearsay are admissible under Fed. R. Evid. 801(d)(2)(E) as statements of

coconspirators only on the condition that a preponderance of the evidence

establishes that a conspiracy existed, that the defendant and declarant were

members, and that the statements were made during the course of and in

furtherance of the conspiracy.”). On occasion, that may also entail a finding

by the judge that a yet-to-be-tried co-defendant was, in fact, a member of

26 the charged conspiracy. 6 Such a finding, as a rule, should be made outside

the presence of the jury. See Tracy,

.

Here, the district court’s comments were made in the context of a

convicted co-defendant’s impending sentencing. In multi-defendant cases,

judges are often called upon to sentence one or more co-defendants while

others are still awaiting trial. Questions of relative culpability may

sometimes be unavoidable, particularly when the defendant being

sentenced claims to have played a lesser role in an overall conspiracy. See

U.S.S.G. § 3B1.2. A judge cannot be said to have manifested partiality

simply by expressing a view of a particular defendant’s culpability based on

information that has been presented to the court That is precisely what the

district court did here.

6 Of course, a co-conspirator statement is admissible if it was made pursuant to any conspiracy, whether charged or uncharged. See United States v. Maldonado-Rivera,

(2d Cir. 1990) (“Though . . . Fed. R. Evid. 801(d)(2)(E) requires proof that both the declarant and the party against whom a declaration is offered be members of the same conspiracy, it does not require that the conspiracy be one charged in the indictment.” (citation omitted)).

27 Third, Wedd attempts to identify partiality in (1) the fact that the

district court spoke with the second jury after declaring a mistrial as to

Wedd and (2) the fact that the district court later received a note from a juror,

with both events assertedly having revealed that the jury’s views largely

favored the Government’s case. We do not follow Wedd’s logic. The district

court’s mere awareness of the jury’s views does not suggest that the district

court adopted such views itself.

Fourth, Wedd points to the district court’s statement, made in

connection with his co-defendant Thompson’s upcoming sentencing, that

there is a high recidivism rate associated with fraud. This innocuous

statement, made while addressing a co-defendant convicted of fraud, had

no bearing on whether the court would be able to preside impartially over

a fraud trial for a separate defendant.

Fifth and finally, Wedd claims that the district court’s evidentiary

rulings created the appearance of partiality. Even where a district court errs

in one or more of its evidentiary rulings, “judicial rulings alone almost never

28 constitute a valid basis for a . . . partiality motion”; flawed rulings are

“[a]lmost invariably . . . proper grounds for appeal, not for recusal.” Liteky,

; see United States v. Colon,

(2d Cir. 1992)

(“[E]arlier adverse rulings, without more, do not provide a reasonable basis

for questioning a judge’s impartiality.”). 7 Moreover, Wedd does not even

identify a flawed aspect of the district court’s evidentiary rulings. While he

nominally challenges the district court’s decision not to admit his travel

records under Federal Rule of Evidence 807, the trial record makes clear that

Wedd failed to provide notice to the Government—which is a requirement

for admission of evidence under Rule 807. See United States v. Morgan,

(2d Cir. 2004) (statement will be admitted under Rule 807 only

if “its proffer follows adequate notice to the adverse party” (internal

quotation marks omitted)). Wedd also cries foul based on the zeal with

7 Cf. S.E.C. v. Razmilovic,

(2d Cir. 2013), as amended (Nov. 26, 2013) (“[A]lthough . . . errors occurred when the court ruled on two motions before they were fully submitted, we see no objective basis for attributing those errors to bias. Judges are human. Errors are made; some are corrected; some are not, but are harmless. Few are attributable to bias.”).

29 which the district court enforced an evidentiary ruling (which, incidentally,

he does not challenge) that neither party alert the jury to the prior trials. But

a court’s careful enforcement of its rulings does not reflect partiality.

Accordingly, we find no error, let alone plain error, in the district

court’s purported failure to recuse itself under Section 455(a).

B. The district court properly instructed the jury on conscious avoidance.

Wedd next challenges the district court’s decision to give a conscious

avoidance jury instruction over his objection. Wedd does not challenge the

charge’s content or wording; rather, he argues that the evidence presented

at trial did not provide a basis for the charge.

The Court reviews “a claim of error in jury instructions de novo,

reversing only where, viewing the charge as a whole, there was a prejudicial

error.” United States v. Ebbers,

(2d Cir. 2006) (internal

quotation marks omitted). “We . . . will not vacate a conviction due to an

erroneous jury charge if the error was harmless.” United States v. Skelos,

, at *5 (2d Cir. Feb. 23, 2021).

30 “The doctrine of conscious avoidance (i.e., willful blindness) prevents

defendants from avoiding criminal liability by deliberately shielding

themselves from clear evidence of critical facts that are strongly suggested

by the circumstances and that, if known, would render them guilty of a

crime.” United States v. Gatto,

(2d Cir. 2021) (internal

quotation marks omitted). “A conscious avoidance jury charge permits a

jury to find that a defendant had culpable knowledge of a fact when the

evidence shows that the defendant intentionally avoided confirming the

fact.”

(internal quotation marks omitted).

A district court may provide a conscious avoidance jury instruction

only if two requirements are satisfied: “(1) the defendant asserts the lack of

some specific aspect of knowledge required for conviction and (2) the

appropriate factual predicate for the charge exists, i.e. the evidence is such

that a rational juror may reach the conclusion beyond a reasonable doubt

that the defendant was [(a)] aware of a high probability of the fact in dispute

and [(b)] consciously avoided confirming that fact.” United States v. Goffer,

31

(2d Cir. 2013) (internal quotation marks and alteration

omitted). Here, the district court properly gave the conscious avoidance

charge because the evidence presented at trial satisfied both requirements.

As to the first requirement of asserted ignorance, Wedd repeatedly

claimed that he was unaware of the auto-subscribing fraud.

As to the second requirement—that the defendant was “[(a)] aware of

a high probability of the fact in dispute and [(b)] consciously avoided

confirming that fact,”

(internal quotation marks omitted)—both

components were satisfied here. The Government presented ample

evidence that Wedd was alerted to the high probability of auto-subscribing,

including: (1) Wedd’s audit determined that Tatto was auto-subscribing; (2)

Wedd knew that Tsvetnenko had previously been involved in auto-

subscribing; (3) Miao directly admitted the fraud to Wedd (as even Wedd

conceded in his testimony); and (4) Wedd received payouts as Tatto and

Tsvetnenko continued to use Mobile Messenger’s platform to auto-

subscribe. Trial evidence likewise supported a conclusion that Wedd took

32 steps to avoid confirming the truth, including that Wedd instructed

potential auto-subscribing partners to coordinate details of the schemes

with his underlings, which a jury could construe as an active attempt to

insulate himself from discussions which he knew would involve direct

acknowledgement of the fraud. 8 Moreover, Wedd never discussed the

results of his internal audit—or Miao’s auto-subscribing admission—with

consumers or law enforcement.

As our precedent shows, in sufficiently suspicious circumstances, a

mere failure to ask questions—even absent more affirmative steps of

avoidance—can demonstrate conscious avoidance. See United States v.

Kozeny,

(2d Cir. 2011) (“[C]onscious avoidance may be

established where[] a defendant’s involvement in the criminal offense may

have been so overwhelmingly suspicious that the defendant’s failure to

8Evidence of this conduct included: (1) Miao’s testimony that Wedd instructed Miao to coordinate details of auto-subscribing with Pajaczkowski; (2) Wedd’s own admission that, even while personally refusing to participate in the auto-subscribing fraud, he told Miao that he could speak with Pajaczkowski; and (3) Wedd’s testimony that he gave Pajaczkowski a great deal of discretion.

33 question the suspicious circumstances establishes the defendant’s

purposeful contrivance to avoid guilty knowledge.” (internal quotation

marks and alteration omitted)). The circumstances in this case were

enormously suspicious, as Wedd was receiving large payouts while a

company that he knew had engaged in auto-subscribing continued to use

his platform. And Wedd’s conduct in this case went beyond a mere failure

to ask questions; as noted above, trial evidence supported a conclusion that

he took active, affirmative steps to avoid confirming the fraud.

Moreover, even if the district court had given the conscious avoidance

charge in error, the error would be harmless in this case. “[A]n unwarranted

conscious avoidance instruction is harmless error where there is

overwhelming evidence of actual knowledge.” United States v. Aina-

Marshall,

(2d Cir. 2003) (internal quotation marks omitted).

Here, the Government presented precisely such overwhelming evidence of

Wedd’s actual knowledge, including financial records tracing payments to

Wedd and detailed, consistent testimony of Wedd’s co-conspirators. For

34 example, Miao testified to having very open conversations with Wedd

about engaging in the auto-subscribing scheme together, including

discussions about how to avoid detection. Pajaczkowski also testified to

speaking with Wedd directly about the schemes, including discussions

about how to keep auto-subscribing partner Tatto under control.

Pajaczkowski further testified to attending a meeting with Wedd and others

as they divided up responsibilities for the Tsvetnenko scheme. And Eromo

testified to openly discussing the Tsvetnenko auto-subscribing scheme with

Wedd. In light of this overwhelming evidence of actual knowledge, even

an erroneous conscious avoidance charge would not warrant a new trial. 9

We therefore reject Wedd’s challenge to the conscious avoidance jury

charge.

9 As we have previously made clear, “[t]he government need not choose between an ‘actual knowledge’ and a ‘conscious avoidance’ theory” and may offer both at trial. United States v. Ferguson,

(2d Cir. 2011) (quoting United States v. Kaplan,

, 128 n.7 (2d Cir. 2007)).

35 C. The Government properly pled and proved aggravated identity theft.

Wedd finally argues that the Government improperly pled

aggravated identity theft in violation of 18 U.S.C. § 1028A(a)(1) and that the

evidence offered at trial was insufficient to support a conviction on Counts

Three and Seven. Although Wedd treats these two issues as largely

indistinguishable, the pleading standard for an indictment is entirely

separate from the evidentiary standard at trial, and therefore we address

each in turn.

1. The Indictment properly alleged aggravated identity theft.

An indictment is sufficient as long as it (1) “contains the elements of

the offense charged and fairly informs a defendant of the charge against

which he must defend,” and (2) “enables [the defendant] to plead an

acquittal or conviction in bar of future prosecutions for the same offense.”

United States v. Alfonso,

(2d Cir. 1998) (internal quotation

marks omitted). “[A]n indictment need do little more than to track the

36 language of the statute charged and state the time and place (in approximate

terms) of the alleged crime.”

Wedd’s Indictment easily meets these basic requirements. Section

1028A(a)(1) provides as follows:

Whoever, during and in relation to any felony violation enumerated in subsection (c), knowingly transfers, possesses, or uses, without lawful authority, a means of identification of another person shall, in addition to the punishment provided for such felony, be sentenced to a term of imprisonment of 2 years.

Correspondingly, Count Three of the Indictment states in relevant part:

From in or about 2011, up to and including in or about 2013, in the Southern District of New York and elsewhere, [Wedd and his co-defendants], during and in relation to a felony enumerated in Title 18, United States Code, Section 1028A(c), to wit, the wire fraud offense alleged in Count Two, did knowingly transfer, possess, and use, without lawful authority, a means of identification of another person, to wit, [Wedd and his co-defendants] used the telephone numbers of consumers without authorization as part of their fraudulent auto- subscribing scheme. (Title 18, United States Code, Sections 1028A & 2.)

App’x at 61-62. Count Seven is similar, varying only with respect to the

dates, the defendants, and the predicate wire fraud offense (Count Six). This

37 language plainly tracks the text of Section 1028A and identifies the

approximate time and place of the alleged crime. See Alfonso,

.

In arguing that the Indictment remains nonetheless inadequate,

Wedd maintains that the specific conduct referenced therein—the auto-

subscribing scheme—did not involve a “use[]” of “a means of

identification” for purposes of Section 1028A. To be sure, the Indictment

does not detail the specific facts of the underlying schemes or precisely how

Wedd and his co-defendants “use[d]” consumers’ telephone numbers. But

there is no requirement that an indictment contain such detail. At the

indictment stage, we do not evaluate the adequacy of the facts to satisfy the

elements of the charged offense. That is something we do after trial.

We have previously explained that “summary judgment does not

exist in federal criminal procedure.” United States v. Sampson,

(2d Cir. 2018). “Unless the government has made what can fairly be

described as a full proffer of the evidence it intends to present at trial . . . ,

38 the sufficiency of the evidence is not appropriately addressed on a pretrial

motion to dismiss an indictment.” Alfonso,

. The district

court must give the Government an opportunity to “make a detailed

presentation of the entirety of the evidence before . . . dismiss[ing] an

indictment on sufficiency grounds,” and the district court lacks the

authority “to require the government, before trial, to make such a

presentation” as this “could effectively force a summary judgment-like

motion on the government.” Sampson,

(internal quotation

marks omitted).

In this case, the Government never purported to make “a full proffer

of the evidence it intend[ed] to present at trial.” Alfonso,

.

Accordingly, there is no merit in Wedd’s argument that the district court

should have looked beyond the four corners of the Indictment to evaluate

the adequacy of its aggravated identity theft pleading. We hold that the

district court properly denied Wedd’s motion to dismiss the Indictment as

to the Section 1028A counts.

39 2. The Government properly proved aggravated identity theft.

After the conclusion of evidence, Wedd moved for a judgment of

acquittal pursuant to Rule 29 of the Federal Rules of Criminal Procedure on

the theory that, because the consumers’ phone numbers originally came to

Mobile Messenger’s database legitimately, subsequent charges resulting

from auto-subscribing could not constitute “use[]” of those numbers for

purposes of Section 1028A. The district court denied the motion, concluding

there was a sufficient basis for the jury to determine that aggravated identity

theft had been proven beyond a reasonable doubt.

We construe this challenge on appeal as one based on sufficiency of

the evidence. This Court reviews de novo “a challenge to the sufficiency of

evidence supporting a criminal conviction,” United States v. Geibel,

(2d Cir. 2004), and will affirm if, ”after viewing the evidence in the

light most favorable to the prosecution, any rational trier of fact could have

found the essential elements of the crime beyond a reasonable doubt,”

Jackson v. Virginia,

(1979); accord United States v. Aquart, 912

(2d Cir. 2018). De novo review similarly applies to “district court

interpretations of . . . federal statutes.” United States v. Henry,

(2d Cir. 2018).

Wedd argues once again that the offense conduct as proved at trial—

that is, a scheme to defraud by means of auto-subscribing unknowing

consumers—cannot constitute a “use[]” of “a means of identification” for

purposes of Section 1028A. Wedd does not dispute that consumers’ phone

numbers—which the fraudsters entered into the auto-subscribing computer

program—are “means of identification,” so the only issue is whether the

fraudsters “use[d]” the numbers within the meaning of the aggravated

identity theft statute. 10

The Government contends that “use[]” ought to be given its ordinary

meaning. See Smith v. United States,

(1993) (“When a word

10See United States v. Dumitru, No. 19-1486-CR,

, at *4 (2d Cir. Mar. 22, 2021) (“We have not had occasion to determine the precise bounds of the aggravated identity theft statute, and we need not do so here because [the defendant’s] actions would fall within those bounds even under a narrow view of the proscribed conduct. “).

41 is not defined by statute, we normally construe it in accord with its ordinary

or natural meaning.”).

We agree. Our analysis begins with the text of the statute. See, e.g.,

United States v. Al Kassar,

(2d Cir. 2011). Section 1028A

proscribes the “knowingly transfer[,] possess[ion,] or use[,] without lawful

authority, [of] a means of identification of another person,” in connection

with certain predicate crimes. 18 U.S.C. § 1028A. In interpreting a statute,

this Court gives “the statutory terms their ordinary or natural meaning.”

United States v. Lockhart,

(2d Cir. 2014) (internal quotation

marks omitted). The dictionary definition of “use” is, in relevant part, “to

put into action or service,” “to avail oneself of,” or to “employ.” See

Definition of Use, MERRIAM WEBSTER, https://www.merriam-

webster.com/dictionary/use; Oxford English Dictionary Online (2020) (“The

act of putting something to work, or employing or applying a thing, for any

. . . purpose”); see also Black's Law Dictionary (11th ed. 2019) (defining “use”

as “[t]o employ for the accomplishment of a purpose; to avail oneself of”).

42 We recently relied on the “ordinary, natural, everyday meaning” of

“use” in defining “use of physical force,” reasoning that “‘use’ requires only

that a person make use of the violent force, convert such force to one’s

service, employ it, avail oneself of it, utilize it, carry out a purpose or action

by means of it, or derive service from it.” United States v. Scott,

, at *8 (2d Cir. Mar. 2, 2021) (en banc) (internal quotation marks and

alterations omitted). We observed that the definition of “use” is

“expansive” and that, “when Congress employs the word ‘use’ in a statute,

its intent is to sweep broadly and not to cabin legislation only to those uses

that most immediately come to mind or that manifest a defendant’s active,

i.e., physical, use.”

(internal quotation marks and alterations omitted).

Drawing upon the ordinary and natural meaning of the statutory text, as

well as our recent decision in Scott, we conclude that to “use” a means of

identification in this setting is to employ or to avail oneself of a means of

identification for a particular purpose.

43 We think the statutory context supports this interpretation of “use”

as it pertains to a “means of identification” for unlawful purposes. Section

1028A(a)(1) criminalizes the knowing and unauthorized use of a means of

identification “during and in relation to” certain enumerated felonies. The

language “during and in relation to” connotes causation. “The salient

point,” as the Sixth Circuit has put it, “is whether the defendant used the

means of identification to further or facilitate the . . . fraud.” United States v.

Michael,

(6th Cir. 2018); see also United States v. Munksgard,

, 1334–35 (11th Cir. 2019) (citing Michael to conclude the same).

Here, the evidence presented at trial proved that Wedd and others

schemed to employ consumers’ phone numbers and, through the auto-

subscribing program, to cause them to pay for text message services without

their knowledge or consent. The victim consumers’ phone numbers plainly

facilitated the fraudulent schemes in which Wedd participated. Put another

way, Wedd employed the phone numbers during and in relation to the

predicate schemes to defraud by means of auto-subscribing. Wedd’s

44 conduct thus falls within this ordinary and natural meaning of the term

“use,” particularly when understood in its statutory context and in the light

of relevant precedent and caselaw. In sum, Wedd unlawfully “use[d]” the

telephone numbers of victims within the meaning of Section 1028A(a)(1). 11

In arguing for a more restrictive definition of the term “use” (the

precise contours of which Wedd does not elucidate), Wedd cites First and

Sixth Circuit decisions concluding that, to “use[]” a person’s “means of

identification” under Section 1028A, a defendant must make some effort to

impersonate the person in question. See United States v. Berroa,

(1st Cir. 2017) (“[W]e read the term ‘use’ to require that the defendant

attempt to pass him or herself off as another person or purport to take some

11Our conclusion that Wedd’s participation in the auto-subscribing schemes falls within the ordinary and natural construction of “use” is consistent with the application of Section 1028A to similar conduct in another case (although the statutory construction was not expressly at issue). See United States v. Greenberg,

(2d Cir. 2016). In Greenberg, the victim customers’ credit card information was retained after they made a purchase and they were charged for joining a frequent shopper club, even though they had never joined such a club or had never received promotion emails or other communications concerning such a club.

at 298–99. The district court’s application of Section 1028A in that case was unquestioned on appeal.

45 other action on another person’s behalf.” (footnote omitted)); United States

v. Miller,

(6th Cir. 2013) (finding “reasonable” the

defendant’s argument that his “conduct does not constitute ‘use’ of [other

people’s] names because he did not steal or possess their identities,

impersonate them or pass himself off as one of them, act on their behalf, or

obtain anything of value in one of their names”(footnote omitted)).

We decline to adopt such a restrictive definition of “use.” As we

explain above, the text, the ordinary and natural meaning, and the statutory

context do not support a construction of “use” that requires impersonation.

Further, insofar as Wedd relies on Miller to argue for this restrictive

definition (upon which the First Circuit’s decision in Berroa likewise relies),

his reliance is misplaced. In Miller, the Sixth Circuit considered whether

“signing a document in [the defendant’s] own name which falsely stated

that [other people] gave him authority, as [an LLC’s] managing member, to

act on behalf of [the LLC] and pledge its property” constituted the “use” of

a means of identification within the meaning of Section 1028A.

.

46 The Sixth Circuit concluded this conduct did not constitute a “use,”

reasoning that “[n]othing inherent in the term ‘uses,’ its placement in the

text of § 1028A, or the statute’s legislative history clearly and definitely

indicates that the term, as applied to the names of persons, is broad enough

to reach the mere act of saying that the persons did something they in fact did not

do.” Id. (emphasis added).

But, unlike the defendant in Miller, Wedd was not charged with the

“mere act of saying that the persons did something they in fact did not do.”

Id. Rather, the schemes to defraud in which Wedd participated involved

enrolling victims, without their consent or knowledge, for paid text message

services by means of auto-subscribing.

Moreover, the Sixth Circuit itself subsequently made clear that Miller

did not establish a requirement “that a defendant impersonate someone

else. . . .” United States v. Michael,

(6th Cir. 2018). In that

later case, the Sixth Circuit gave “use” a “fairly straightforward”

construction, as we have done here, holding that “[t]o ‘use’ a means of

47 identification in this setting is ‘[t]o convert to one’s service’ or ‘to employ’

the means of identification.”

at 626 (citing a number of dictionaries). 12 In

so holding, the Sixth Circuit expressly rejected the defendant’s argument

that “that the statutory object of the sentence, using a ‘means of

identification’ for fraudulent purposes, . . . confines the coverage of the law

only to impersonations.”

.

Accordingly, insofar as Wedd’s challenge to his convictions for

Counts Three and Seven is based on the sufficiency of the evidence, we reject

it. There was more than sufficient evidence for a rational factfinder to

conclude that Wedd was aware of, facilitated, and participated in the auto-

subscribing schemes and that he, in fact, received the proceeds from those

schemes over several years. In short, we hold that there was sufficient

12As the Sixth Circuit went on to explain, the language of Section 1028A “no doubt covers impersonations, and impersonations may well have been one of the targets, perhaps even the principal target, of this sentencing-enhancement statute. But it is not unusual for the words of laws to go beyond the central, even the sole, motivation for enacting them.” Michael,

. “[O]nly the words of a law, not the motivations of its authors, may cabin (or for that matter extend) its reach.”

48 evidence for a rational jury to find Wedd guilty beyond a reasonable doubt

of violating Section 1028A.

Wedd argues that, in the alternative, if his conduct constitutes a

“use[]” under § 1028A, then the statute is unconstitutionally vague. But

Wedd’s case does not require the Court to define the outer limits of the term

“use[],” as Wedd’s conduct falls squarely within both the plain dictionary

definition of the term. And “a plaintiff who engages in some conduct that

is clearly proscribed cannot complain of the vagueness of the law as applied

to the conduct of others.” Holder v. Humanitarian Law Project,

, 18-

19 (2010) (internal quotation marks and alteration omitted).

Accordingly, we hold that the district court properly denied Wedd’s

motion to enter a judgment of acquittal on the Section 1028A counts.

III. Conclusion

In sum, we hold as follows:

1. The district court did not err, much less plainly err, in not recusing

itself from Wedd’s retrial pursuant to

(a). The district court’s

brief comment on Wedd’s relative culpability was based on information it

49 learned during the course of proceedings and therefore did not create an

appearance of partiality.

2. The district court properly instructed the jury on conscious

avoidance because sufficient trial evidence supported that theory of

criminal liability.

3. The district court properly denied the motion to dismiss Counts

Three and Seven. The Indictment adequately pled a violation of 18 U.S.C.

§ 1028A because it tracked the statutory text of the charged offense and

stated the approximate time and place of the alleged crime.

4. The district court properly denied the motion for a judgment of

acquittal on Counts Three and Seven. There was sufficient evidence for a

rational jury to conclude beyond a reasonable doubt that Wedd violated 18

U.S.C. § 1028A because, by employing victims’ phone numbers during and

in relation to the auto-subscribing schemes to subscribe them to text

message services without their knowledge or consent, Wedd “use[d],

without lawful authority, a means of identification of another person.”

50 We therefore AFFIRM the judgment of the district court.

51