Deponent’s Use of A.I. to Answer Deposition Questions Barred; ChatGPT Was Not an Attorney

Why It Matters

The ruling establishes that AI tools are not a shield for unrepresented parties, limiting AI‑driven self‑representation and shaping discovery strategy for pro se litigants.

Key Takeaways

• Court barred deponent from using ChatGPT during deposition
• Privilege claim rejected because deponent had no attorney
• Case dismissed with prejudice for repeated discovery failures
• Ruling signals courts will limit AI assistance for pro se litigants
• Similar AI concerns appear in other 2026 deposition cases

Pulse Analysis

The rapid diffusion of generative AI, especially large language models like ChatGPT, has reshaped how parties gather and process legal information. Pro se litigants, often lacking resources, have turned to these tools for instant research, document drafting, and even real‑time guidance during proceedings. Courts have welcomed efficiency gains but remain wary of the technology’s potential to undermine procedural safeguards, create undisclosed assistance, and blur the line between self‑representation and unauthorized practice of law.

In the Michigan case, the court confronted a deponent who openly consulted ChatGPT while answering deposition questions. Judge [redacted] determined that the AI platform could not be covered by attorney‑client privilege because no attorney was present, and that the deponent’s refusal to answer on that basis violated discovery obligations. After multiple extensions and warnings, the court exercised its sanctioning authority under Federal Rules 37(b)(2) and 41(b) to dismiss the case with prejudice, emphasizing that willful non‑compliance outweighs any perceived benefit from AI assistance.

The precedent sends a clear signal to litigants and counsel: AI may be a valuable research aid, but it cannot replace legal counsel during sworn testimony. Law firms are now advised to issue explicit protocols prohibiting AI use in depositions and to monitor remote proceedings for unauthorized tools. As courts continue to grapple with AI’s role, future rulings may refine the boundaries between permissible research and impermissible real‑time assistance, shaping the evolving landscape of e‑discovery and courtroom technology.