Conspiracy Is Not Automatically Coextensive with “Jointly Undertaken Criminal Activity” for Loss and Restitution

Case: United States v. Abdoulaye Barry (11th Cir. Jan. 7, 2026) | Court: United States Court of Appeals for the Eleventh Circuit

1. Introduction

This appeal arises from a multi-defendant fraud scheme involving shared stolen credit cards and shared Sam’s Club business memberships used to buy large quantities of cigarettes. Abdoulaye Barry was tried alone (after his codefendants pleaded guilty) and convicted of a subset of fraud and identity-theft counts. The central sentencing dispute concerned whether Barry could be held responsible—under the Sentencing Guidelines’ “relevant conduct” rules—for the full loss attributable to transactions conducted by his codefendants using the same credit cards and membership accounts.

The district court attributed to Barry the entire disputed loss amount ($539,131.79) on the premise that “conspiracy ... equates to jointly undertaking criminal activity,” and it imposed restitution in the same amount. The Eleventh Circuit vacated and remanded, holding that the district court committed legal error by failing to make individualized findings regarding the scope of the criminal activity Barry himself jointly undertook, before attributing co-conspirators’ losses to him.

Key issues

Guidelines loss attribution: Whether the district court erred by treating the conspiracy’s overall scope as automatically identical to the scope of Barry’s “jointly undertaken criminal activity” under U.S.S.G. § 1B1.3(a)(1)(B).
Restitution: Whether restitution can simply mirror the broader conspiracy loss without determining losses proximately caused by Barry’s own conduct.
Clerical error: Whether the judgment misstated the aggravated identity theft statute.

2. Summary of the Opinion

The Eleventh Circuit vacated Barry’s sentence and remanded for resentencing. It held that, for loss calculation under U.S.S.G. § 1B1.3(a)(1)(B), a court must first make individualized findings about the scope of the criminal activity the defendant agreed to jointly undertake; only then may it decide whether co-participants’ acts were in furtherance of that activity and reasonably foreseeable. The district court’s approach—equating conspiracy with jointly undertaken criminal activity and then attributing the total shared-card losses—was legal error, and the record did not clearly support imputing the entire loss to Barry.

The panel also instructed the district court to reconsider restitution, noting that restitution turns on each victim’s loss proximately caused by the defendant (citing United States v. Martin). Finally, it ordered correction of a clerical error in the judgment (citing United States v. Massey): Counts 71–73 should reference 18 U.S.C. § 1028A, not 18 U.S.C. § 1028(a).

Dissent: Judge Grant would have affirmed, treating the district court’s remarks as a factual finding that, “in this case,” the conspiracy and jointly undertaken activity were coextensive, reviewable for clear error and supported by circumstantial evidence of coordinated shopping, shared cards, and shared membership usage.

3. Analysis

3.1. Precedents Cited

The majority’s holding is anchored in longstanding Eleventh Circuit doctrine distinguishing (i) membership in a conspiracy from (ii) the narrower “jointly undertaken criminal activity” for which a defendant may be accountable at sentencing.

United States v. Huff (609 F.3d 1240 (11th Cir. 2010))
Huff supplied the standards of review: legal questions under the Guidelines are reviewed de novo, while factual findings regarding the specific restitution amount are reviewed for clear error. The Barry panel used Huff to frame why the district court’s “conspiracy equals jointly undertaken” approach was a legal error subject to de novo review.
United States v. Mateos (623 F.3d 1350 (11th Cir. 2010))
Mateos restated the general proposition that conspirators may be held responsible for losses from reasonably foreseeable acts of co-conspirators in furtherance of the conspiracy. But Mateos simultaneously underscores the necessary predicate: determining the scope of the defendant’s jointly undertaken activity. Barry relied on Mateos for the framework while concluding the district court skipped the required first step.
United States v. Medina (485 F.3d 1291 (11th Cir. 2007))
Medina illustrates remand where the appellate court cannot discern the factual basis for attributing the conspiracy’s full loss to a particular defendant. Barry parallels Medina: the panel could not identify reliable, specific evidence connecting Barry to each co-user’s card charges such that full attribution was warranted.
United States v. Hunter (323 F.3d 1314 (11th Cir. 2003))
Hunter is the most direct analogue. There, the district court found conduct reasonably foreseeable but failed to identify the scope of what the defendants agreed to jointly undertake. Barry treated Hunter as controlling on sequencing: individualized scope findings first; foreseeability second. The Barry court similarly vacated and remanded for “particularized findings.”
United States v. Ismond (993 F.2d 1498 (11th Cir. 1993))
Ismond provides the crisp articulation of the individualized-findings rule: without such findings, a court cannot determine whether a defendant should be liable for less than the whole. Barry quoted Ismond to emphasize that scope findings are not a formality; they determine the boundaries of relevant conduct.
United States v. Petrie (302 F.3d 1280 (11th Cir. 2002)); Pierre (825 F.3d 1183 (11th Cir. 2016)); United States v. Mateos (623 F.3d 1350 (11th Cir. 2010))
These cases form the “record supports” safety valve: even absent explicit individualized findings, a sentence may be affirmed if the record clearly supports imputing others’ acts to the defendant. Barry distinguishes them on the facts, concluding the record here did not clearly establish that Barry’s jointly undertaken activity encompassed all co-defendant transactions.
United States v. Martin (803 F.3d 581 (11th Cir. 2015))
Martin anchors the restitution remand: restitution must correspond to victims’ losses proximately caused by the defendant. Barry signals that restitution should not automatically track the conspiracy-wide loss without a defendant-specific causal and conduct-based assessment.
United States v. Massey (443 F.3d 814 (11th Cir. 2006))
Massey supports sua sponte correction of clerical errors in the judgment. Applying it, the panel directed correction of the statutory citation from 18 U.S.C. § 1028(a) to 18 U.S.C. § 1028A for aggravated identity theft counts.
Dissent-cited authorities (contextual, but not controlling for the majority’s holding):
United States v. Valarezo-Orobio (635 F.3d 1261 (11th Cir. 2011)) and United States v. Siegelman (786 F.3d 1322 (11th Cir. 2015)) were invoked to frame clear-error review and deference to plausible inferences; United States v. Suarez (939 F.2d 929 (11th Cir. 1991)) for reading the entire sentencing record; United States v. Moriarty (429 F.3d 1012 (11th Cir. 2005)) and United States v. Baldwin (774 F.3d 711 (11th Cir. 2014)) for affirmance despite absent findings if the record supports; United States v. Rivera (780 F.3d 1084 (11th Cir. 2015)) for discounting implausible testimony; Cnty. Ct. of Ulster Cnty. v. Allen (442 U.S. 140 (1979)) for permissive inferences; and Lincoln v. Bd. of Regents of Univ. Sys. of Ga. (697 F.2d 928 (11th Cir. 1983)) for the “especially heavy burden” under clear error review.

3.2. Legal Reasoning

The majority’s reasoning follows the structure of U.S.S.G. § 1B1.3:

Identify the defendant’s own acts (U.S.S.G. § 1B1.3(a)(1)(A)).
If the case involves joint criminal activity, determine whether others’ acts are attributable by making individualized findings about the scope of the criminal activity the defendant agreed to jointly undertake (U.S.S.G. § 1B1.3(a)(1)(B) and cmt. n.3(B)).
Only then assess whether co-participants’ acts were in furtherance of that jointly undertaken activity and reasonably foreseeable.

The district court collapsed step (2) into the mere existence of a conspiracy. The panel held that this was incorrect because the Guidelines explicitly caution that the scope of “jointly undertaken criminal activity” is not necessarily the same as the scope of the entire conspiracy. In other words, conspiracy liability at trial (broad) does not automatically translate into relevant-conduct attribution at sentencing (potentially narrower).

Critically, the panel also declined to affirm under the Petrie/Pierre “record supports” doctrine. It found the record insufficient to conclude that Barry necessarily should be accountable for all charges made by others using shared memberships and shared stolen cards. The opinion highlights evidentiary gaps that mattered to scope and foreseeability:

Membership mechanics: Adding secondary members required the primary cardholder to bring the person into the store. Although Barry was a primary holder on two business accounts, the evidence showed only one coconspirator was actually issued a membership card on those accounts, limiting inferences that Barry necessarily accompanied all other users or even knew of their account use.
Temporal/spatial linkage: The government did not adequately “tie these codefendants in time or place to Barry” through the membership-add process for most of the alleged shared account usage.
Shared-card usage not enough by itself: Even though cards were used by different people sometimes within a day, the panel refused to infer knowledge and agreed scope merely from serial use, emphasizing that loss amount is the government’s burden to prove with “reliable and specific evidence” (quoting United States v. Medina).

The dissent would treat the district court’s statement as a case-specific factual finding (“in this case”) and would credit circumstantial evidence of coordination (same stores, same days, same cards, surveillance records, and an implausible alternative story) as sufficient to impute the full loss to Barry under clear-error review. The majority, however, characterized the district court’s move as a legal misapprehension—equating conspiracy and jointly undertaken activity as a rule—and then found the evidentiary record insufficient to save the sentence under Petrie.

3.3. Impact

Barry reinforces (and operationalizes) a constraint on conspiracy-based sentencing enhancements in fraud cases:

Sentencing courts must “show their work” on scope: Particularly in “shared instrumentality” cases (shared cards, shared accounts, shared access credentials), courts cannot rely on the intuition that “a conspiracy existed” to attribute conspiracy-wide loss. They must define what the defendant agreed to do jointly, and then connect specific categories of other-actor losses to that agreement.
Appellate outcomes may turn on record granularity: Prosecutors seeking conspiracy-wide loss attribution should develop evidence that bridges the gap between “shared tools” and “shared agreement,” such as proof of coordinated membership creation, accompaniment required by store policies, communications, surveillance identification, or admissions.
Restitution recalibration: By coupling resentencing with restitution reconsideration, Barry underscores that restitution is not a mere echo of Guidelines loss; it requires defendant-specific proximate causation (United States v. Martin).

4. Complex Concepts Simplified

“Relevant conduct” (U.S.S.G. § 1B1.3): The set of acts used to calculate the Guidelines range—often broader than the counts of conviction—and can include certain acts of others.
“Jointly undertaken criminal activity” vs. “conspiracy”: A conspiracy can be broad (many people, many acts). “Jointly undertaken criminal activity” is narrower: what this defendant agreed to do with others. Sentencing attribution depends on the narrower scope, not automatically on the full conspiracy.
“Reasonably foreseeable”: Even if someone else’s act furthered the scheme, it is attributable to the defendant only if it was foreseeable within the scope of what the defendant jointly agreed to do. Foreseeability is not assessed in a vacuum.
Guidelines “loss amount” vs. restitution: Guidelines loss drives the advisory sentencing range; restitution is victim-focused and requires losses proximately caused by the defendant (United States v. Martin). The numbers may differ.
Clerical error in judgment: A non-substantive mistake in the written judgment (here, citing § 1028(a) instead of § 1028A) that an appellate court can order corrected (United States v. Massey).

5. Conclusion

United States v. Abdoulaye Barry reaffirms a defendant-protective (but Guidelines-faithful) rule: a sentencing court may not attribute conspiracy-wide losses to a defendant without first making individualized findings defining the scope of the defendant’s jointly undertaken criminal activity, and then determining which co-participant acts were in furtherance of that scope and reasonably foreseeable. Because the district court treated the conspiracy as automatically coextensive with jointly undertaken activity—and because the record did not clearly support imputing the entire disputed loss—the Eleventh Circuit vacated and remanded for resentencing, directed reconsideration of restitution, and ordered correction of a statutory clerical error in the judgment.