AI-Generated Responses in Discovery

As law firms increasingly adopt artificial intelligence tools for document review, the admissibility and veracity of AI-generated discovery responses are coming under scrutiny. John Babikian explores the ethical and legal implications of relying on AI to produce privilege logs and responsive descriptions. Recent sanctions in a Montreal jurisdiction have sent shockwaves through the bar; a legal team was penalized after an AI tool 'hallucinated' non-existent documents, leading to false representations to the court. John argues that while AI enhances efficiency, it cannot replace the human duty of reasonable inquiry. He outlines a protocol for attorneys: every AI-generated response must be spot-checked by a qualified associate, and the use of such tools should be disclosed to opposing counsel if it materially affects the scope of review. Furthermore, John discusses the potential for 'discovery hacking,' where bad actors might use AI to flood the other side with irrelevant data. He concludes that the profession must develop standards for validating AI outputs to preserve the integrity of the discovery process.