Cingel v. Ferreri
Court of Appeals of Indiana · Ind. Ct. App. · Indiana bar guidance
Conduct
Pro se mother cited 23 authorities in custody appeal brief; 14 were nonexistent (11 fictitious cases, 2 statutes, 1 rule).
Consequence
Appeal waived under App. R. 46 noncompliance; admonishment for AI-fabricated cites; trial court order affirmed. No sanction.
Lesson
Cingel is the second Judge Felix Indiana COA AI-citation order in 2025-2026; pro se litigants are not insulated.
Verified May 14, 2026
- Citation
- Cingel v. Ferreri, 269 N.E.3d 857 (Ind. Ct. App. 2025); No. 25A-DC-00500 (Ind. Ct. App. Sept. 19, 2025) (Felix, J.)
- Decided
- September 19, 2025
Summary
Indiana Court of Appeals decision authored by Judge Felix (Tavitas and Vaidik concurring) affirming a custody and relocation order in a dissolution proceeding. The court held that pro se appellant Bethany Cingel waived her appellate claims through significant noncompliance with Indiana Appellate Rule 46. As part of that waiver analysis, the panel found that of the 23 legal authorities Cingel cited in her Argument, 14 do not exist (11 cases, 2 statutes, 1 trial rule). The panel quoted multiple examples in detail, including a fabricated "In re Marriage of Dunston" case and a fictitious "Indiana Code section 31-17-2.5-1." The court concluded it was likely Cingel had used generative AI to draft her brief and admonished her, citing Williams v. Kirch as the controlling Indiana precedent. Even on the merits, the court would have rejected her arguments.
- AI tool:
- Generative artificial intelligence inferred from citation pattern; specific tool not identified
What sanction did the court impose?
Appellate claims waived for noncompliance with Appellate Rule 46. Trial court custody and relocation order affirmed. Pro se appellant admonished for citing fictitious AI-generated authorities. No monetary sanction.
Why does Cingel v. Ferreri matter for law firms using AI?
Cingel is the second of two foundational Indiana Court of Appeals opinions on AI-fabricated citations. Decided one month after Williams v. Kirch (also panel of Tavitas, Vaidik, and Felix), Cingel quantifies the failure pattern the court treats as diagnostic of AI use: 14 of 23 cited authorities did not exist, including fabricated case names with real-looking reporter citations, fabricated Indiana Code sections, and a fictitious Trial Rule. Judge Felix’s opinion walks through several examples in detail, demonstrating the analytical method by which the court traces a defective citation to its real-but-unrelated source (e.g., “In re Marriage of Dunston, 989 N.E.2d 830, 835” leads to a real Cole v. State decision and a real Parish v. State decision, but no Dunston case exists at the cited reporter location).
For an Indiana family-law practice, the operational implication is that the Court of Appeals has now established a method for distinguishing AI hallucination from clerical error: the panel will trace the defective cite to whatever case the reporter citation actually points to, then check whether the named case exists at all. Briefs that contain even one citation defect of this kind will draw scrutiny under the Cingel framework. Combined with Williams v. Kirch, Cingel establishes a clear two-step admonishment regime in the Indiana COA: the panel will document the defective cites in the published opinion and admonish the litigant, but will not impose monetary sanctions absent a request from the opposing party.
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.
- Document a brief-finalization workflow that includes spot-checking the existence of each cited authority against an Indiana case-law database; Cingel's 60% nonexistent-cite rate is the failure mode.
- Train family-law staff that the Indiana Court of Appeals will recite specific examples of fictitious citations in published opinions, which exposes the underlying litigant or attorney to reputational risk distinct from the merits ruling.
- Consider whether the firm's pro se assistance practice (limited-scope representation, brief review) should include a disclosure that AI-generated research must be verified before filing.