Christopher A. Ambrose v. Bandy X. Lee
U.S. District Court, District of Connecticut · D. Conn. · Connecticut bar guidance
Conduct
Pro se defendant in a defamation case filed motions to dismiss containing serious misreadings of cited authorities; the court suspected AI-assisted drafting and warned but did not sanction.
Consequence
No-tolerance policy declared prospectively; court reserved sanctions for future filings (AI-assisted or not) that hallucinate or severely misstate the law without reasonable excuse.
Lesson
Courts can issue prospective AI-no-tolerance warnings based on suspicion from substantive case misreadings; an AI admission is not required.
Verified May 10, 2026
- Citation
- Ambrose v. Lee, No. 3:25-cv-00398 (SVN), Mem. & Order (D. Conn. Mar. 31, 2026), ECF No. 50; follow-on dismissal order at ECF No. 54 (D. Conn. Apr. 15, 2026)
- Decided
- March 31, 2026
Summary
Christopher A. Ambrose, an attorney representing himself, filed a defamation and related-tort suit against Dr. Bandy X. Lee in the District of Connecticut. Defendant Lee, also self-represented, moved to dismiss (ECF No. 30) and later filed a supplemental motion to dismiss (ECF No. 47) alleging that Ambrose had obtained in forma pauperis status by misrepresenting his finances. On March 31, 2026, Judge Sarala V. Nagala issued a 36-page Memorandum and Order (ECF No. 50) granting Lee's initial motion to dismiss in part as to one count and denying it in all other respects. In two footnotes within the order (note 4 on page 9 and a parallel note on page 26), the Court flagged that it 'suspects' Lee used artificial intelligence to draft portions of her brief, based on 'serious misreading' of cited authorities. On April 15, 2026, after reviewing the financial submissions, the Court issued a follow-on order (ECF No. 54) granting in part Lee's supplemental motion under 28 U.S.C. § 1915(e)(2)(A) and dismissing the action without prejudice for an untrue allegation of poverty.
- AI tool:
- Unspecified generative AI
- Sanction amount:
- None (warning only; the Court declared a no-tolerance policy for AI-assisted briefing that hallucinates legal propositions or severely misstates the law, and stated such filings 'will often result in sanctions absent reasonable excuse')
What sanction did the court impose?
Judge Nagala did not impose sanctions on Lee for the suspected AI use in the March 31 order; instead, the Court issued a no-tolerance warning that applies prospectively to all parties: 'all parties are on notice that the Court has a no-tolerance policy for any briefing (AI-assisted or not) that hallucinates legal propositions or otherwise severely misstates the law. Such filings will often result in sanctions absent reasonable excuse.' The Court cited Willis v. U.S. Bank Nat'l Ass'n, No. 3:25-CV-516-BN, 2025 WL 1408897 (N.D. Tex. May 15, 2025) as authority for Rule 11 application to AI-assisted briefing by pro se litigants. The Court reiterated this warning later in the order: 'Given the serious misreading of these cases, the Court again suspects Defendant used AI to draft this portion of her brief as well. The Court reiterates its no tolerance policy for the use of AI in drafting pleadings.' On the merits, Count Five was dismissed and all other counts survived; the April 15, 2026 follow-on order then dismissed the case without prejudice under 28 U.S.C. § 1915(e)(2)(A) on a separate ground (Ambrose's untrue IFP affidavit, which omitted a substantial Fidelity IRA portfolio and Writers Guild residual royalties).
Why does Christopher A. Ambrose v. Bandy X. Lee matter for law firms using AI?
The procedural posture of this case is unusual: the AI-relevant content is not in a sanctions order or a holding, but in two footnotes within a substantive motion-to-dismiss ruling. Footnote 4 on page 9, attached to the Court’s discussion of Lee’s reliance on Tyler, reads: “Given Defendant’s serious misreading of Tyler, the Court suspects she used artificial intelligence (‘AI’) to draft at least that portion of her brief. The Court cautions Defendant that the use, without verification, of the accuracy of the information AI platforms generate, like any other shoddy research method from other sources or tools, implicates Federal Rule of Civil Procedure 11.” Page 26 carries a parallel note: “Given the serious misreading of these cases, the Court again suspects Defendant used AI to draft this portion of her brief as well. The Court reiterates its no tolerance policy for the use of AI in drafting pleadings.”
The court’s structural move matters. Nagala did not require an admission from Lee, did not hold an evidentiary hearing on AI use, and did not impose sanctions. The “suspicion” framing plus the prospective no-tolerance statement together create a doctrinal posture that lower courts in the Second Circuit can borrow: the court can announce a Rule-11-anchored AI policy without needing the procedural predicate that an actual finding would require. Willis v. U.S. Bank, 2025 WL 1408897 (N.D. Tex.) is the cross-circuit cite Nagala used to anchor the rule’s application to pro se litigants, which is significant for defendants and plaintiffs alike who proceed without counsel and rely on AI-assisted research.
For partners thinking about how this fits the broader sanctions corpus, Ambrose v. Lee is distinct from Med. Buyers Group v. Pence (M.D. Ga.), Johnson v. Dunn (N.D. Ala.), or Benjamin v. Costco (E.D.N.Y.). Those cases all involve a court finding bad faith and imposing monetary sanctions after the offending party admitted to AI use. Ambrose v. Lee sits at the upstream procedural moment: a court spotting the AI pattern in a substantive ruling, declining to sanction now, but setting up the docket for sanctions on any future filing that repeats the conduct. That upstream posture is increasingly common in pro se AI matters and reads as the emerging “first occurrence” judicial response in 2025-2026 federal practice.
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.
- Track Nagala's no-tolerance framing as the D. Conn. authority on Rule 11 and AI-assisted briefing for pro se litigants; the warning is a prospective policy statement, not a one-off finding, and Willis v. U.S. Bank (N.D. Tex.) is the cross-circuit cite Nagala used to anchor the rule.
- Recognize the structural move: the Court did not need to find AI use as a matter of fact to issue the warning; suspicion based on serious misreading of cases was sufficient. Defense counsel against pro se litigants suspected of AI use can point to this case for the proposition that a court can warn (and later sanction) without an admission.
- Treat the 'misreading of Tyler' / 'misreading of these cases' framing as the operational signal for AI suspicion: courts are pattern-matching on substantive misuse of cited authority, not on stylistic or formatting tells.
- Document for any firm filing pro se on its own behalf (e.g., partner defending against malpractice claim) that the lawyer-pro-se rule (Plaintiff is an attorney, 'not entitled to the same solicitude as a regular pro se litigant') applies in the Second Circuit context as articulated by Nagala.