Chumpitaz-Morales v. Bondi
U.S. Court of Appeals for the Tenth Circuit · 10th Cir. · Colorado bar guidance , Kansas bar guidance , New Mexico bar guidance , Oklahoma bar guidance , Utah bar guidance , Wyoming bar guidance
Conduct
Pro se petitioner's BIA appeal brief misrepresented the holdings of several cited cases; panel attributed to AI hallucination.
Consequence
BIA denial affirmed; sanctions expressly declined (government did not request them); written reminder to all litigants.
Lesson
10th Cir. treats holdings-misrepresentation (not just fabricated cites) as AI failure mode; sanction posture tracks government's request.
Verified May 8, 2026
- Citation
- Chumpitaz-Morales v. Bondi, No. 25-9527 (10th Cir. Feb. 11, 2026) (Carson, Baldock, Kelly, JJ., op. by Baldock, J.) (unpublished)
- Decided
- February 11, 2026
Summary
Pro se petitioners Flora Chumpitaz-Morales and her minor daughter D.S.T.C., natives of Peru, sought review of a Board of Immigration Appeals order denying asylum, withholding of removal, and CAT relief. Petitioners had been represented before the BIA but proceeded pro se before the immigration judge and on appeal to the Tenth Circuit. The panel (Carson, Baldock, Kelly) determined that Chumpitaz-Morales's brief "misrepresents the holdings of several cases it cites" and attributed the misrepresentations to "an artificial intelligence (AI) product (e.g., ChatGPT, Microsoft CoPilot) which 'hallucinated' the cases' purported holdings." The government had identified one misrepresented citation in its response brief; the court treated the misrepresentations as a pattern but did not enumerate the full list of misused authorities.
- AI tool:
- Generative AI inferred (panel: 'an artificial intelligence (AI) product (e.g., ChatGPT, Microsoft CoPilot)')
What sanction did the court impose?
Petition for review denied on the merits (no protected-ground nexus established for the Peru extortion claim). The court expressly declined to sanction Chumpitaz-Morales, observing that the government did not request sanctions, and instead included a written reminder that "all litigants, represented and unrepresented, must read their filings and take reasonable care to avoid misrepresentations, factual and legal."
Why does Chumpitaz-Morales v. Bondi matter for law firms using AI?
Chumpitaz-Morales sits at the most lenient rung of the Tenth Circuit’s emerging AI-citation remedy ladder: misrepresentation found, sanction expressly declined, written reminder to all litigants. The panel’s reasoning ties the lenience to two factors. First, the government did not request sanctions; the Tenth Circuit explicitly calibrated remedy to the adversarial posture of the case rather than acting on its own inherent authority. Second, the misrepresentation went to holdings rather than to citation existence: the cited cases were real, the problem was that Chumpitaz-Morales’s brief described their holdings in ways the actual opinions did not support. The court treats this as an AI-hallucination failure mode distinct from fabricated citations, naming both ChatGPT and Microsoft CoPilot as exemplar tools.
For institutional defendants in immigration matters, the operational point is that requesting sanctions matters. The same Carson/Baldock/Kelly panel issued a warning in Dodds two days earlier on the same date without a sanctions request from the State (Dodds was a habeas matter where the State did not affirmatively seek Rule 11-equivalent consequences). The Tenth Circuit’s pattern across this chambers cluster is to grant whatever sanction the prevailing party asks for, calibrated down to a written reminder where no request is made. Firms should treat the sanctions request itself as a strategic decision in any brief responding to AI-hallucinated opposition; the request frames the panel’s remedy ceiling.
Implications for your firm
Operational steps a firm reading this case may wish to consider documenting. Strategic and rule-application calls belong to your firm's attorneys.
- Holdings-misrepresentation, not just fabricated citations, is now an enumerated AI-hallucination failure mode in the 10th Cir.; firms reviewing pro se opposition briefs should flag misused authorities even when the citation itself resolves to a real case.
- The panel's express observation that 'the government did not request sanctions' is doctrinally meaningful: the Tenth Circuit calibrates sanction posture in part to whether the prevailing party seeks them. In adversarial postures, requesting Rule 38 or chambers-rule sanctions is the action that triggers the analysis.
- Carson/Baldock/Kelly is now a three-case AI-warning cluster (Chumpitaz, Dodds, Biglow); firms can map AI-citation expectations to this panel composition for petitions for review and other 10th Cir. matters.
Sources
Primary sources
- The opinion identifies one misrepresented citation in the government's response brief but does not enumerate the full list of misused authorities; identifying every case Chumpitaz-Morales misrepresented would require comparing the petitioner's brief against the panel's analysis.
- No Westlaw assignment is visible in the unpublished opinion's PDF header at this verification pass.