A Law Firm's AI-Assisted Brief Cited a Nonexistent Case, and the Court Refused to Let Them Fix It
What happened
Boies Schiller Flexner is one of the most prominent litigation firms in the United States. When the firm filed a respondents' brief in Bixler v. Church of Scientology International in July 2025, the work was reportedly prepared with AI assistance. What opposing counsel found when they checked the citations was not what anyone at the firm intended to submit.
Counsel for the other side identified a series of material citation errors. Authorities cited in the brief were mischaracterized, titles were wrong, and at least one case could not be found in the legal record at all: it did not exist. In federal litigation, citation accuracy is not a cosmetic concern. Judges and clerks rely on cited authorities to be real, findable, and accurately described. A brief that cites a phantom case is not merely sloppy; it is a submission that tells the court it should not be trusted.
Partner John Kucera accepted responsibility for the failure to verify the citations. The firm moved to replace the brief with a corrected version. The court denied that request. A ruling that rejects a substitution is the court making something explicit: the filing is on the record, the opposing party has already read it, and a quiet swap after the fact is not how accountability works. The court later moved toward considering monetary sanctions against the firm.
This is not the first time AI-assisted legal filings have surfaced fabricated citations. Courts across the country have confronted similar problems since large-language-model tools became common in legal practice, and lower-profile practitioners have faced fines for the same conduct. What distinguishes this matter is the institutional context. Boies Schiller's profile makes it harder to attribute the failure to resource constraints or inexperience with professional obligations. If the verification gap exists at that level of practice, the same gap almost certainly runs deeper through the profession.
The accountability problem here extends beyond whether someone checked before filing. What the court has no way to assess is what the tool actually produced, which portions of the brief came from it, what the review process looked like, and whether any lawyer compared the output against real sources before signing. Without that record, the firm can offer only an apology and a motion. A provable record of what a system generated, what a practitioner reviewed, and when, would make that chain visible at the moment it matters rather than after opposing counsel has already caught the error.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.