Use of AI backfires on an expert witness in court, tell me about this news in (https://www.msn.com/en-us/technology/artificial-intelligence/use-of-ai-backfires-on-an-expert-witness-in-court/ar-AA2cwytP?ocid=BingNewsVerp)

The news report covers a multimillion-dollar lawsuit in Texas involving victims of a factory explosion, in which an expert witness faced intense scrutiny after relying heavily on ChatGPT to draft his testimony report.

Here are the key details of the incident:

• The Incident and Subject: The expert witness, Jason Autenrieth (director of reliability and operations at KnightHawk Engineering), was retained by corporate manufacturing defendant 3M. Plaintiffs’ attorneys discovered that Autenrieth used ChatGPT to generate more than 80% of his expert report.

• What the Prompts Revealed: Opposing counsel uncovered roughly 300 pages of prompt logs submitted to ChatGPT. The prompts went beyond simple proofreading; they included basic inquiries asking what specific industrial machines were and how they functioned, as well as prompts directly asking the AI how to attack and undermine the claims of opposing witnesses.

• The Defense: Autenrieth and his legal representation acknowledged using AI to assist with drafting, structuring, and formatting, but maintained that he conducted his own independent analysis across thousands of pages of records. They argued that he did not delegate his core professional judgment or conclusions to the model.

• Broader Legal Implications: Legal scholars and trial experts warned that generative AI models are inherently sycophanticdesigned to please the user and reflect whatever bias or premise is fed into the promptwhich directly clashes with an expert witness’s legal duty to maintain objective, independent judgment. The case highlights growing concerns in courtrooms over automated bias, hallucinations, and compromised credibility when AI is substituted for professional expertise.