If You Use A.I. for Litigation, You Need Know About This New Case Law ASAP

Why It Matters

The decision instantly removes a layer of confidentiality for AI‑assisted legal work, forcing law firms to rethink client communications and risk management, and signaling broader regulatory scrutiny of AI in litigation.

Key Takeaways

• New SDNY ruling strips AI queries of attorney‑client privilege
• AI‑generated case data now discoverable in criminal and civil suits
• Clients risk self‑incrimination by feeding confidential facts to AI
• Courts view unsupervised AI use as unethical and potentially malpractice
• Lawyers must supervise AI tools; avoid client‑driven prompts without counsel

Summary

A federal judge in the Southern District of New York issued a landmark ruling that any information a party inputs into an artificial‑intelligence system for legal advice is not shielded by attorney‑client privilege. The decision, handed down by Judge Raymond Rayoff in late February, marks the first time a court has explicitly addressed AI‑generated communications in the context of privilege.

The ruling treats AI prompts as ordinary documents that can be subject to discovery, whether in criminal or civil matters. In the case at hand, a defendant fed factual details into the Claude AI model before a grand‑jury indictment, and the prosecution was allowed to seize those inputs as evidence. The court also rejected the notion that such exchanges constitute work‑product, effectively opening a new frontier for litigants to expose confidential strategy.

Vince White described the outcome as “terrifying,” while Michael Hilfrey warned that using AI is a “damned if you do, damned if you don’t” scenario. He cited two recent client incidents: one where AI was used to polish language in a low‑stakes veterinary bill dispute, and another where AI‑generated answers were later questioned at a deposition, resulting in adverse consequences for the client.

Practitioners now face a clear mandate to supervise any AI usage and to counsel clients against entering case‑specific data into generative tools without attorney oversight. Failure to do so could expose privileged information, trigger discovery disputes, and even raise ethical or malpractice concerns, reshaping how the legal industry adopts emerging technology.

Original Description

I actively track the comments on these videos and try to respond within 24 hours, so please feel free to reach out. However, be careful not to share identifying information about yourself on this publicly viewable forum. Consider hitting the like and subscribe buttons or even just commenting below. Doing so helps me help more people like you.

If you have a question about employment law, workplace discrimination, or sexual harassment, please do not hesitate to contact us. We will try to craft a video response for you, or where more appropriate, we are happy to have a private conversation.

At our firm, we believe everybody works - but not everybody wins. You need to be smart out there.

The Merch: https://www.bonfire.com/store/nyc-job-attorney/

Learn more about our firm here: https://www.NYCJobAttorney.com

See our cases and results in the press here: https://www.nycjobattorney.com/press/

Some jurisdictions may consider this attorney advertising, although advertising is not the intent of these videos.

This answer does not constitute legal advice and you should contact an attorney to confirm or research further any statements made in this answer. Any statements of fact or law I have made in this answer pertain solely to Federal law as construed within the EEOC or the laws of New York State and should not be relied upon in any way in any other jurisdiction.