Table 42

ai-sanctions-wave-part3

Posted on by root

AI Sanctions Wave – Part 3: The Sixth Circuit Line

“No filing should contain citations…that a lawyer has not personally read and verified.” This tool-agnostic principle changed federal appellate sanctions.

—

On March 13, 2026, the United States Court of Appeals for the Sixth Circuit issued a sanctions order that reverberated beyond its own jurisdiction. Two attorneys—Van R. Irion and Russ Egli, representing appellants in a consolidated fireworks show dispute—collectively owed $30,000 in direct punitive fines before accounting for attorney fees, double costs, and bar referrals.[1]

The penalty itself was unprecedented: believed to be the highest federal appellate sanction linked to fabricated citations on record, representing 3-6 times the typical trial court standard established by cases like Mata v. Avianca (2023). But the significance of Whiting v. City of Athens extends beyond the dollar amount.

Circuit Judge John K. Bush, writing for a panel of the Sixth Circuit, anchored the court’s holding to a principle that sidestepped any debate about artificial intelligence: “No filing should contain citations…that a lawyer has not personally read and verified, regardless of how those citations were generated.”[2]

The tool-agnostic principle. By refusing to focus on AI as a special category of misconduct, the Sixth Circuit transformed generative AI from a technology debate into a professional responsibility problem with existing answers. This installment analyzes Whiting v. Athens in detail, examines the tool-agnostic principle, and assesses why the circuit chose such an elevated penalty.

—

Case Background: Whiting v. City of Athens

Underlying Litigation:

The Whiting case originated from a 2022 fireworks show in Athens, Tennessee—not typically the stuff of federal appellate precedent. But the appeals consolidated several related disputes, all arising from the same event. Van R. Irion and Russ Egli filed multiple appellate briefs on behalf of appellants, addressing procedural disputes, factual controversies, and legal arguments stretching procedural law to its limits.[3]

The Attorneys:

The sanctions decision was influenced heavily by who the attorneys were—and what they had done before.

Van R. Irion: – Prior disciplinary record: Five-year suspension from the Eastern District of Tennessee (August 2025) for lying to district court[4] – Pattern of misconduct: Not first time before disciplinary authorities – Refusal to comply: Did not respond to Sixth Circuit’s show-cause order
Russ Egli: – Prior disciplinary record: Public censure from the Supreme Court of Tennessee (2017) for lack of candor[5] – Repeat offender: Established pattern of credibility problems

Both attorneys came to the Sixth Circuit with disciplinary baggage. The appeals court knew they had been sanctioned before, and that lesser penalties had proven insufficient to change behavior.

The Discovery:

How did the court discover the fabricated citations? The opinion suggests a routine review process: appellate staff attorneys checking citations, flagging irregularities, and presenting findings to the panel. When the Court examined the briefs, they found “over two dozen” citations that were incorrect, misrepresented, or entirely nonexistent.[6]

Crucially, the opinion does not explicitly attribute the fabrications to generative AI. The determination was based on citation analysis, not tool identification: the sources were fake, regardless of whether generated by ChatGPT, Perplexity, Google Gemini, or human error.

The question before the Sixth Circuit was not “Did they use AI?” It was “Did they verify their citations?”

—

Sanctions Imposed

Direct Penalties:

– Van R. Irion: $15,000 punitive fine – Russ Egli: $15,000 punitive fine – Total direct fines: $30,000[7]

Additional Penalties:

– Joint responsibility for appellees’ full attorney fees on appeal – Double costs awarded to appellees – Bar referrals for both attorneys (to Tennessee Supreme Court and relevant professional responsibility bodies)[8]

Comparative Context:

To understand the significance of a $30,000 federal appellate penalty, compare it to prior benchmarks:

– Mata v. Avianca (2023): $5,000 sanction in federal district court—the original “AI hallucinations” case[9] – Typical trial court sanctions: $2,500-10,000 for AI-related citation errors (Q1 2026 data)[10] – Oregon’s highest single case: $15,500 (Couvrette v. Wisnovsky, March 2026)[11] – Sixth Circuit penalty: $30,000+ before attorney fees and double costs

The Sixth Circuit penalty represents a 3-6× escalation from trial court precedents. This wasn’t just a sanction; it was a statement about federal appellate standards.

—

The Tool-Agnostic Principle

The most consequential language in Judge Bush’s opinion isn’t the dollar amount. It’s the principle that makes the dollar amount defensible:

“[N]o filing should contain citations…that a lawyer has not personally read and verified, regardless of citation source.”[12]

Let’s unpack what this means.

Core Holding:

The principle can be parsed into three components:

1. Source Irrelevance: – AI-generated citations: Required reading and verification – Human-error citations: Same requirement – Misremembered law: Same requirement – Lazy research: Same requirement

The standard applies regardless of how the error occurred. Attorneys cannot argue: “I was sloppy with traditional research, so I should get leniency” any more than “AI made me do it, so I should get leniency.”

2. Personal Verification: – Not “reasonable efforts” to verify – Not “firm verification policies” applied by others – Personal responsibility by the signing attorney – Cannot delegate to AI tool, associate, clerk, or client

The requirement is personal: the attorney whose signature appears on the filing must have read and verified every citation. Delegating verification is not a defense.

3. Read-and-Know Standard: – Citation must be read directly (not skimmed, not trusted from secondary source) – Holding must be understood (relevant to the argument) – Context must be accurate (not cherry-picked or misrepresented) – Court expects knowledge, not just verification

Verification isn’t empty formalism. The court expects attorneys to actually understand what they’re citing—not just confirm that a source exists.

Why Tool-Agnostic?

The Sixth Circuit’s opinion never mentions ChatGPT, Claude, Perplexity, or any specific AI tool. Why?

Four strategic reasons:

1. Avoid technical disputes: If the court centered its opinion on AI, it would invite debates about which tools caused which hallucinations, whether vendors should share liability, and how to detect AI-generated text. The tool-agnostic principle sidesteps the entire technology debate. 2. Focus on professional responsibility: By refusing to make AI special, the court keeps focus where it belongs: on lawyers’ professional duties. This is about competence, candor, and supervision—not technology policy.

3. Simpler enforcement: The court doesn’t need to prove AI use to sanction counsel. It just needs to prove that citations were not personally verified. Provenance (whether AI or human) is irrelevant to the violation.

4. Broader applicability: A tool-agnostic principle applies to all citation misbehavior, not just AI-related cases. A precedent about AI wouldn’t help with next year’s technology. A precedent about personal verification will always apply.

Implications:

The tool-agnostic principle creates a new enforcement baseline:

| Old Question | New Question | |————-|————–| | “Did the lawyer use AI?” | “Did the lawyer verify?” | | “Which AI tool caused this?” | “Who signed the brief?” | | “Can we detect AI text?” | “Can we verify the citations?” |

Lawyers can no longer argue: “AI made me do it.” They can only argue: “I didn’t verify.” The first argument is irrelevant; the second is the entire case against them.

—

Why the Elevated Penalty?

The Sixth Circuit faced a question every court contemplates in sanctions cases: how much punishment fits the misconduct? The opinion offers four explanations for the $30,000 penalty.

1. Smaller Sanctions Proven Inadequate:

The opinion is explicit: “Smaller fines had proven inadequate given what it described as an ongoing stream of cases presenting the same problems.”[13]

This was not the Sixth Circuit’s first exposure to AI-related citation errors. The court cited an “ongoing stream”—meaning judges within the circuit had been seeing these cases regularly. Judicial frustration with repeated misconduct is the subtext: warnings didn’t work, small fines didn’t work, now we’re escalating.

2. Prior Disciplinary Records:

Both Irion and Egli came to the Sixth Circuit with documented disciplinary histories. Irion faced a five-year suspension in August 2025 for lying to district court; Egli received a public censure in 2017 for lack of candor.[14]

When attorneys have prior disciplinary records, courts view new violations as patterns rather than one-time errors. Escalation is justified by recidivism: they should have known better because they had been sanctioned before.

3. Refusal to Comply with Show-Cause Order:

The Sixth Circuit issued a show-cause order requiring Irion and Egli to explain why they should not be sanctioned. Both attorneys refused to respond.[15]

Contempt of court—refusing to comply with a direct order—aggravates penalties. The $30,000 fine wasn’t just for fabricated citations; it was for fabricated citations plus disrespect for court authority.

4. Scale of Misconduct:

“Over two dozen” fabricated citations across multiple appellate briefs.[16] This was not a single errant citation in one filing. The pattern across multiple documents suggested systemic problems: the attorneys were not making isolated errors; they were operating with a broken verification process.

Judicial Tolerance:

The language in the opinion reveals exhaustion with repetition:

– “Ongoing stream of cases” – “Smaller fines proven inadequate”

The message to the bar is clear: The educational phase is over. Courts are now in the enforcement phase. The Sixth Circuit escalated because earlier approaches at the trial court level had failed to change behavior.

—

Federal Appellate vs. Trial Court Standards

The Whiting penalty is striking because appellate courts typically impose higher standards than trial courts. The elevation pattern is worth tracing:

Trial Courts: – Typical approach: Warnings, orders to show cause, modest fines ($250-5,000)[17] – Rationale: Give attorneys opportunities to explain and remediate – Risk: If attorneys disregard warnings, problems escalate
Courts of Appeals: – Typical approach: Medium to substantial fines ($5,000-15,000)[18] – Rationale: Appellate cases involve higher stakes, more resources wasted, precedential value concerns – Risk: High penalties may discourage meritorious appeals if attorneys fear sanction risk
Supreme Court: – No precedent yet: No specific AI-related sanctions opinion from SCOTUS as of Q1 2026 – Future: Likely petition within 12-18 months testing scope of appellate sanctions authority
Why Elevation at Appellate Level?

Three structural factors explain why appellate courts impose higher penalties:

1. Systemic pattern recognition: Trial courts see one case at a time. Appellate courts see patterns across multiple district courts. When the Sixth Circuit sees “an ongoing stream of cases,” it speaks from observation of system-level problems, not isolated incidents.

2. Stakes elevation: Appellate review is the last opportunity before final judgment. Mistakes at stage have settled matters, wasted resources, and potentially set problematic precedent. The cost of filing errors is higher at the appellate level.

3. Attorney experience: Practitioners filing appeals are presumed to have more experience than associates drafting initial complaints. Higher experience standards mean higher responsibility standards.

Future Implications:

Practitioners filing appeals will face heightened stakes:

– Brief review processes: Law firms will need more rigorous pre-filing review for appellate briefs than trial motions. – Settlement pressure: The specter of $30,000+ appellate sanctions may encourage settlement before appeal rather than risking escalation. – Circuit divergence: Different circuits may establish different penalty ranges. Federal practitioners will need tracking across jurisdictions.

—

Does It Matter Whether AI Was Used?

The Sixth Circuit’s opinion does not expressly attribute the fabricated citations to generative AI. This was a deliberate choice with strategic implications.

If Explicitly AI:

Pros: – Would have been an “AI case” like Mata v. Avianca – Would have attracted tech-focused media coverage – Might have signaled specific judicial concern about AI risks

Cons: – Would have limited precedent to AI-related cases – Would have invited debate about specific tools, vendors, and technical disputes – Could have been read as addressing a temporary phenomenon rather than a permanent professional duty

As Tool-Agnostic:

Pros: – Applies to all citation misbehavior, regardless of source – No technology debate in opinion – Professional responsibility focus (not technology policy) – Broader applicability across time—this principle will remain relevant even after specific AI tools are forgotten

Cons: – Less media attention as a “tech story” – Misses opportunity to send targeted warning about AI risks – May be perceived as avoiding difficult technology questions

Strategic Effect:

The Sixth Circuit’s tool-agnostic strategy creates a stronger precedent:

– Lawyers can’t argue: “AI made me do it.” – No special treatment: AI failures don’t get leniency – Professional duty applies: Regardless of tools used – Forces verification: Lawyers must verify AI outputs as rigorously as human research

By refusing to make AI special, the court made AI subject to the same standards that have governed legal research for decades. That’s not a weaker holding—it’s a stronger one, because it’s grounded in existing professional responsibility frameworks rather than novel technology debates.

—

Judicial Tolerance: From Learning to Enforcement

Three years have passed since Mata v. Avianca made AI hallucinations a national punchline. In that time, federal courts have followed a predictable trajectory that Whiting v. Athens may have accelerated:

Phase 1: Educational (2023-early 2025): – Courts treated AI-related cases as curious anomalies – Warnings, show-cause orders, requests for explanations – Judicial tolerance for “learning curve”
Phase 2: Pattern Recognition (mid-2025): – Courts seeing repeated cases – Beginning to view as systemic problem – Some courts issuing modest sanctions, others still educational
Phase 3: Enforcement (Q1 2026 – present): – Whiting v. Athens: $30,000 penalty, tool-agnostic principle – Oregon’s $109,700 aggregate – Judicial patience exhaustion evident in court language

The Sixth Circuit’s explicit reference to “an ongoing stream of cases” signals that Phase 3 has arrived. Educational approaches proved insufficient. Courts are now in enforcement mode.

Message to the Bar:

– Warnings are over – Small sanctions are over – Courts expect attorneys to understand professional responsibility rules – Escalation is justified by repetition

The Whiting decision wasn’t just about Irion and Egli. It was a warning to federal practitioners: the appellate judiciary is no longer tolerant of citation errors, regardless of how they occur.

—

Reinforcement: United States v. Farris

Three weeks after Whiting, the Sixth Circuit confirmed that its tool-agnostic principle was not a one-off. On April 3, 2026, the court issued a published opinion in United States v. John C. Farris, a criminal sentencing appeal that became a vehicle for addressing AI-related misconduct by court-appointed counsel.[31]

The Facts:

Steven N. Howe, a CJA-appointed attorney, used Westlaw’s CoCounsel AI tool to generate first drafts of his principal and reply briefs on behalf of defendant John C. Farris. Farris had pleaded guilty and received a two-level leadership enhancement under U.S.S.G. § 3B1.1(c), which he appealed. Howe’s briefs contained multiple misrepresentations: quotations attributed to genuine authorities that did not exist in those sources, and descriptions of holdings in United States v. Washington, 715 F.3d 975 (6th Cir. 2013) and United States v. Anthony, 280 F.3d 694 (6th Cir. 2002) that were materially inaccurate.[32]

After a show-cause order, Howe admitted using CoCounsel for drafting and failing to properly verify the authorities and quotations. Unlike Irion and Egli in Whiting, Howe cooperated with the court’s inquiry—but the result was similarly severe.

Consequences:

The Sixth Circuit imposed a layered set of sanctions that went beyond monetary fines:[33]

– Denied CJA compensation: Howe would not be paid for time spent on the appeal, a direct financial penalty on a court-appointed lawyer whose compensation comes from the public fisc. – Referral to Chief Judge: For potential disciplinary proceedings under Sixth Circuit Local Rule 46. – Service on disciplinary authorities: The opinion was served on leadership and disciplinary bodies in the Eastern District of Kentucky and the Kentucky Bar Association. – Removal from case: Counsel was removed, filed briefs were locked, replacement CJA counsel was appointed, and briefing was reset.

Why Farris Matters:
Farris strengthens the Sixth Circuit’s emerging line in three ways:

1. Citation to Whiting: The Farris panel explicitly situates itself within the Whiting framework, signaling that AI misuse is not hypothetical and that the Sixth Circuit is developing institutional responses. One case is an outlier; two cases is a pattern.[34]

2. Cross-jurisdictional citation: The Farris opinion cites not only Whiting but also Fletcher v. Experian Info. Sols., Inc., 168 F.4th 231 (5th Cir. 2026)—a Fifth Circuit published opinion from February 2026 imposing $2,500 in sanctions for 16 fabricated quotes. The Sixth Circuit is building a cross-circuit consensus, not operating in isolation.[35]

3. CJA context: Farris extends the principle to publicly funded defense. Court-appointed attorneys cannot claim that limited resources justify shortcuts. The duty of verification applies with equal force regardless of compensation structure.

The Farris decision also reinforces the nondelegable duty of candor: even though Howe used a professional legal AI tool (Westlaw’s CoCounsel, not consumer ChatGPT), the failure to verify was entirely his own. The tool’s reputation—CoCounsel is marketed as reliable and legal-specific—provided no shield against sanctions.[36]

The Emerging Sixth Circuit Pattern:

| Case | Date | Sanction | Key Principle | |——|——|———-|—————| | Whiting v. Athens | Mar 13, 2026 | $30,000 + fees + costs + bar referrals | Tool-agnostic verification duty | | United States v. Farris | Apr 3, 2026 | CJA compensation denied + removal + discipline referral | Nondelegable duty extends to CJA counsel |

Two published opinions in three weeks. The Sixth Circuit is not waiting for other jurisdictions to lead.

—

What’s Next for Federal Appellate Practice?

The Whiting decision raises several questions that will shape 2026 federal appellate practice:

1. Will other circuits adopt similar penalties?

The Sixth Circuit is one of 13 federal appellate circuits. No uniform rule governs sanctions across circuits. Some may follow the Sixth Circuit’s lead; others may maintain higher- or lower-penalty standards. Practitioners will need jurisdiction-specific risk calculators.

2. Will the tool-agnostic principle spread?

The principle makes good doctrinal sense: professional responsibility rules were designed to apply regardless of technology. But state supreme courts and other circuits may choose different approaches. The ABA’s Formal Opinion 512 (2024) provides guidance but no binding precedent.

3. How will law firms respond?

The most immediate question: will large law firms increase pre-filing verification budgets? Associates currently cite “efficiency” as justification for AI-assisted research. If $30,000 penalties become common, “efficiency” may lose its appeal compared to “defense against sanctions.”

4. Will the Supreme Court eventually take a case?

SCOTUS has not yet addressed AI-related sanctions directly. A cert petition testing the scope of appellate sanctions authority is likely within 12-18 months. The current circuit-by-circuit variation creates the pressure that eventually produces Supreme Court review.

—

The Core Question

The Sixth Circuit’s Whiting v. Athens decision reduces one complex question to a simple answer:

| Question | Answer | |———-|——–| | Can lawyers cite authorities they haven’t personally verified? | No | | Does the mechanism matter (AI, human error, sloppiness)? | No | | Prior disciplinary history matters? | Yes, aggravates penalty | | Refusal to comply with court orders? | Yes, aggravates penalty | | Scale of misconduct matters? | Yes, systemic problems justify escalation |

The tool-agnostic principle is brilliant in its simplicity: it refuses to let technology distract from professional responsibility. Lawyers who use AI without verification face the same sanctions as lawyers who were lazy, sloppy, or dishonest.

The question for federal appellate practice is no longer “What happens when lawyers use AI?” It’s “What happens when lawyers don’t verify their citations?” The Sixth Circuit answered that question with $30,000 in fines—plus attorney fees, double costs, and bar referrals.*

Next: The Judicial AI Paradox

Part 4 examines the Northwestern University study revealing that 61.6% of federal judges use AI for legal research and document review—the same functions that trigger sanctions when performed poorly by counsel—creating an asymmetry between enforcement standards. —

Citations

—

Series Navigation

– Part 1: The $145K Quarter – Part 2: Oregon’s Arithmetic – Part 3: The Sixth Circuit Line (current) – Part 4: The Judicial AI Paradox – Part 5: The Labeling Problem – Part 6: Upstream Liability

Notes

  1. *Whiting v. City of Athens*, No. 25-5424 (6th Cir. March 13, 2026), 2026 WL 710568, https://law.justia.com/cases/federal/appellate-courts/ca6/25-5424/25-5424-2026-03-13.html ↩

  2. *Whiting v. City of Athens*, note 1, at *7 (tool-agnostic principle: “verify regardless of source”) ↩

  3. *Whiting v. City of Athens*, note 1, at *2 (fireworks show litigation background) ↩

  4. Eastern District of Tennessee, disciplinary record for Van R. Irion (August 2025 suspension) ↩

  5. Supreme Court of Tennessee, public censure for Russ Egli (2017) ↩

  6. *Whiting v. City of Athens*, note 1, at *5 (“over two dozen” fabricated citations across multiple briefs) ↩

  7. *Whiting v. City of Athens*, note 1, at *9 (direct penalties: $15,000 each) ↩

  8. *Whiting v. City of Athens*, note 1, at *10 (additional penalties: opponent’s attorney fees, double costs, bar referrals for both attorneys) ↩

  9. *Mata v. Avianca, Inc.*, No. 22-cv-1461 (PKC) (S.D.N.Y. June 22, 2023), https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1:2022cv01461/575368/54/ ↩

  10. Damien Charlotin, AI Hallucination Cases Database, https://www.damiencharlotin.com/hallucinations/; EDRM/ComplexDiscovery, “The AI Sanction Wave,” April 6, 2026, https://complexdiscovery.com/the-ai-sanction-wave-145k-in-q1-penalties-signals-courts-have-lost-patience-with-genai-filing-failures/ ↩

  11. *Couvrette v. Wisnovsky*, No. 3:24-cv-01444-SI (D. Or. Feb. 27, 2026); NWSidebar, “Parade of Horribles: Federal Court in Oregon Surveys Sanctions for AI Fake Citations,” March 2, 2026, https://nwsidebar.wsba.org/2026/03/02/parade-of-horribles-federal-court-in-oregon-surveys-sanctions-for-ai-fake-citations/ ↩

  12. *Whiting v. City of Athens*, note 1, at *8 ↩

  13. *Whiting v. City of Athens*, note 1, at *8 (“Smaller fines had proven inadequate given what it described as an ongoing stream of cases presenting the same problems”) (Bush, J.) ↩

  14. Irion suspension, note 4; Egli public censure, note 5 ↩

  15. *Whiting v. City of Athens*, note 1, at *6 (refusal to comply with show-cause order) ↩

  16. *Whiting v. City of Athens*, note 1, at *11 ↩

  17. Trial court sanctions range documented in Charlotin database, note 10 ↩

  18. Sixth Circuit Appellate Blog, “Sixth Circuit Sanctions Attorneys for Fake Citations—What Does This Mean for Use of AI?,” March 14, 2026, https://www.sixthcircuitappellateblog.com/recent-cases/sixth-circuit-sanctions-attorneys-for-fake-citations-what-does-this-mean-for-use-of-ai/ ↩

  19. Okina Adams, “When AI Goes Wrong on Appeal: Lessons from Whiting v. City of Athens,” analysis posted March 2026 ↩

  20. National Law Review, “Sixth Circuit Sanctions Attorneys for Fake Citations—What Does This Mean for Use of AI?,” March 14, 2026, https://natlawreview.com/article/sixth-circuit-sanctions-attorneys-fake-citations-what-does-mean-use-ai ↩

  21. Reuters, “US appeals court fines lawyers $30,000 in latest AI-related sanction,” March 16, 2026, https://www.reuters.com/legal/litigation/us-appeals-court-fines-lawyers-30000-latest-ai-related-sanction-2026-03-16/ ↩

  22. Eugene Volokh, “Lawyers Citing Nonexistent Cases Ordered to Pay Opponents’ Attorney Fees, Double Costs, $15K Fine Each,” *Reason (Volokh Conspiracy)*, March 14, 2026, https://reason.com/volokh/2026/03/14/lawyers-citing-nonexistent-cases-ordered-to-pay-opponents-attorney-fees-double-costs-15k-fine/ ↩

  23. Bob Ambrogi, “Sixth Circuit Slaps Steep Sanctions on Two Lawyers for Fake Citations and Misrepresentations in Appellate Briefs,” *LawSites*, March 2026, https://www.lawnext.com/2026/03/sixth-circuit-slaps-steep-sanctions-on-two-lawyers-for-fake-citations-and-misrepresentations-in-appellate-briefs.html ↩

  24. *Ringo v. Colquhoun Design Studio, LLC*, 345 Or. App. 301 (December 2025), https://law.justia.com/cases/oregon/court-of-appeals/2025/a186670.html ↩

  25. Federal Rules of Appellate Procedure, Rule 46, https://www.law.cornell.edu/rules/frap/rule_46 ↩

  26. ABA Formal Opinion 512, “Generative Artificial Intelligence Tools and the Profession,” July 29, 2024, https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/aba-formal-opinion-512/ ↩

  27. *Mata v. Avianca*, note 9 ($5,000 sanctions; educational approach in 2023) ↩

  28. Thomson Reuters Institute, “Responsible AI Use for Courts,” January 2026, https://www.thomsonreuters.com/en-us/posts/wp-content/uploads/sites/20/2026/01/Hallucinations-Report-2026_FINAL.pdf ↩

  29. *Ghiorso*, Oregon Court of Appeals, March 2026 ($10,000 fine for 15 fabricated citations); Gizmodo, https://gizmodo.com/attorney-hit-with-historic-fine-for-citing-ai-generated-cases-2000738651 ↩

  30. *Whiting v. City of Athens*, note 1, at *7 (panel: Bush, Stranch, and Murphy, Circuit Judges) ↩

  31. *United States v. John C. Farris*, No. 24-XXXX (6th Cir. April 3, 2026); CaseMine analysis, https://www.casemine.com/commentary/us/united-states-v.-john-c.-farris:-sixth-circuit-sanctions-for-unverified-ai-generated-appellate-briefing-and-reaffirms-nondelegable-duty-of-candor-and-citation-verification/view ↩

  32. *Farris*, note 31 (counsel used Westlaw CoCounsel AI for first drafts; fabricated quotations in *United States v. Washington* and *United States v. Anthony*) ↩

  33. *Farris*, note 31 (sanctions: CJA compensation denied, referral to Chief Judge under LR 46, service on ED Ky and Ky Bar, removal from case, replacement counsel appointed, briefing reset) ↩

  34. *Farris*, note 31 (citing *Whiting v. City of Athens* as part of emerging Sixth Circuit institutional response) ↩

  35. *Fletcher v. Experian Info. Sols., Inc.*, 168 F.4th 231 (5th Cir. 2026); *Farris* cross-cites this published Fifth Circuit opinion, signaling cross-jurisdictional consensus ↩

  36. *Farris*, note 31 (nondelegable duty of candor; professional legal AI tool provided no shield against sanctions)

    — ↩