:

SULLIVAN & CROMWELL BLAMES AI FOR COURT FILING ERRORS

AI DESK1 MIN READ
WED, APR 22, 2026

■ AI-SUMMARIZED FROM 1 SOURCE BELOW

Elite Wall Street law firm Sullivan & Cromwell apologized to a New York federal judge after admitting that AI hallucinations caused errors in a major case filing. The firm's global restructuring co-head Andrew Dietderich disclosed the mistakes in a letter to Judge Martin Glenn on Saturday.

Sullivan & Cromwell acknowledged that artificial intelligence generated false information in documents submitted for the Prince Group case. The errors represent a significant stumble for one of the country's most prestigious law firms. Dietderich's letter to Judge Glenn came after the firm discovered the hallucinations—instances where AI systems generated plausible-sounding but entirely fabricated information. The specific nature and scope of the errors were not detailed in available reports. The incident highlights growing risks as law firms integrate AI tools into their workflows without adequate safeguards. Legal professionals are increasingly relying on AI for document review and research, but these systems can produce confident-sounding false citations and information. Sullivan & Cromwell's admission comes amid broader industry concerns about AI reliability in high-stakes legal work. The firm did not specify what corrective measures it would implement, though the apology suggests the court may require remedial filings or other action to address the false information in the record.

■ SOURCES

The Guardian — Technology

■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE

■ MORE FROM THE AI DESK

Anthropic's Head of Growth Amol Avasare indicated the company's current Pro and Max subscription tiers no longer align with how users actually deploy Claude. The hint emerged after Anthropic briefly removed Claude Code from Pro for new customers, then reversed the decision following backlash.

1H AGOAI Desk

Google is integrating generative AI capabilities into Maps, expanding the platform's functionality beyond navigation. The enhancement aims to deliver smarter, more intuitive features to billions of users worldwide.

2H AGOAI Desk

Pangram Labs' AI detection Chrome extension has flagged content warning about artificial intelligence as itself being AI-generated, raising questions about the tool's accuracy and the authenticity of papal statements circulating online.

5H AGOAI Desk

AI-generated content is undermining civilian safety in active conflict areas, according to researcher Rachel Adams. The threat highlights disparities in how Western tech companies address disinformation globally.

5H AGOIndustry Desk

■ SUBSCRIBE TO THE DAILY BRIEF

ONE EMAIL, 5 STORIES, 06:00 UTC. UNSUBSCRIBE ANYTIME.