The prominent law firm Sullivan & Cromwell has formally admitted to the U.S. Bankruptcy Court that it submitted documents containing "hallucinated" citations generated by artificial intelligence. The incident occurred during the legal proceedings for Prince Group, a case closely watched by the financial and legal sectors for its potential implications on corporate restructuring and digital asset oversight. This admission highlights the growing risks of integrating generative AI into high-stakes legal environments without rigorous human oversight.
Failure of Internal AI Verification Policies
According to court filings from April 9, 2026, the law firm utilized AI tools that produced fabricated and distorted legal precedents, including non-existent statutes. Sullivan & Cromwell representatives stated that the error was a result of failing to adhere to an internal policy summarized as
"trust nothing, verify everything." This policy requires all attorneys to undergo specific training before using AI-assisted research tools and mandates the manual verification of every output.
Opposition Identifies 28 Fabricated Citations
The discrepancies were brought to the court's attention by the opposing counsel, Boies Schiller Flexner. The firm identified a significant number of errors in the original motion, including:
- At least 28 incorrect citations ranging from misattributed rulings to entirely fictional case law.
- Direct quotes from court documents that do not exist in any official records.
- Distorted interpretations of existing statutes that favored the filer's position.
Following these revelations, Boies Schiller Flexner requested a postponement of the scheduled hearing to review the validity of the revised documents. Sullivan & Cromwell has since withdrawn the compromised filing and submitted a corrected version to the court.
Impact on Legal Tech and Digital Asset Proceedings
As blockchain-based companies and cryptocurrency entities increasingly navigate the bankruptcy process, the reliance on automated legal tools has come under scrutiny. This incident serves as a cautionary tale for the broader fintech and legal industries regarding the reliability of LLMs (Large Language Models) in providing factual data. While AI can streamline administrative tasks, the Prince Group case demonstrates that the "hallucination" effect remains a critical vulnerability in the professional application of the technology.
Frequently Asked Questions
Quick answers to the most common questions about this topic.
For educational purposes only. Nothing here constitutes financial or investment advice. Crypto markets are highly volatile — always DYOR before making any decisions.
Fact-checked per our editorial policy. We maintain strict independence from advertisers. Spot an error? Let us know.