In a Q&A published in the Practical Law Journal, a publication of Reuters, on Oct. 1, 2025, Judge Paul W. Grimm (ret.), director of the Bolch Judicial Institute of Duke Law, explains the evidentiary challenges courts face with the rise of AI.
He distinguishes between AI-generated evidence, created entirely by generative AI tools, and AI-enhanced evidence, where human- or computer-produced content is modified by AI software. Both, he notes, raise critical questions of authentication, reliability, and potential prejudice at trial.
Grimm points to proposed Federal Rule of Evidence 707, which would require courts to apply the rigorous standards of Rule 702 to machine-generated evidence, ensuring judges act as gatekeepers to protect juries from unreliable or misleading AI content.