TL;DR

A New York federal judge has terminated a case after attorney Steven Feldman repeatedly submitted court filings containing fake AI-generated citations. One filing stood out for its “conspicuously florid prose,” including a Ray Bradbury quote and gardening metaphors.

When AI Goes Wrong in Court

District Judge Katherine Polk Failla took the unusual step of terminating a case this week after an attorney’s persistent misuse of AI in court filings. Attorney Steven Feldman had been asked multiple times to correct filings that contained fabricated citations, but kept responding with documents that included yet more fake references.

One filing particularly caught the judge’s attention for its strikingly different style — packed with what Failla called “conspicuously florid prose” that contrasted with the grammatical errors and run-on sentences in Feldman’s other submissions.

Bradbury Quotes and Biblical Passages

The problematic filing included an extended quote from Ray Bradbury’s Fahrenheit 451 and metaphors “comparing legal advocacy to gardening and the leaving of indelible ‘mark[s] upon the clay.’” Another passage that raised the court’s eyebrows invoked a Bible passage about divine judgment as an apparent means of acknowledging the lawyer’s failure to catch the fake citations.

These “out-of-left-field” references to ancient libraries and literary works made the AI-generated nature of the text plainly visible to the court.

A Growing Pattern

This case adds to a growing list of incidents where lawyers have faced consequences for submitting unverified AI-generated content. The sanctions highlight the gap between how quickly legal professionals are adopting AI tools and how well they understand the technology’s limitations — particularly its tendency to generate plausible-sounding but entirely fabricated case citations.

For law firms and solo practitioners, the message from the bench is becoming harder to ignore: AI output demands the same rigorous verification as any other source of legal research.