3M expert used ChatGPT to draft '0% at fault' defense report
TL;DR
- The 350 pages of ChatGPT session logs entered into evidence surfaced during a three-hour deposition recess after plaintiff's counsel spotted a five-page AI-generated 'Citation Overlay' document.
- ChatGPT itself flagged the '0% responsible' framing as an easy target for opposing counsel when the expert used it to run mock cross-examination prompts against his own report.
- KnightHawk Engineering publicly called coverage of the incident fake news and characterized AI as a legitimate professional tool, a position no bar association has formally endorsed.
A Harris County jury awarded $61 million to plaintiffs suing 3M over the January 2020 Watson Grinding explosion in Houston and assigned 30 percent of the fault to the company. The trial record now shows that 3M's own expert witness generated most of his report using ChatGPT.
Court records reviewed by 404 Media show plaintiffs' attorney Will Moye discovered a five-page 'Citation Overlay' during discovery and forced production of roughly 350 pages of ChatGPT prompts from Josh Autenrieth of Knighthawk Engineering, the firm 3M paid roughly $90,000 for its analysis. Autenrieth billed at $475 an hour. Among the prompts he fed the chatbot: 'create an exceptional expert witness report defending the standard of care at 3M' and 'show how 3M is 0% at fault for the explosion at Watson Grinding.' Trial testimony pegged the report at roughly 85 to 90 percent ChatGPT output.
"This expert relied on AI not as an assistive device, but exclusively relied on ChatGPT to form his opinions and write his report," Moye told 404 Media.
Autenrieth defended the work from the stand. He testified that "my opinions were put in there, and AI helped me to draft a straw man to build off of," that he altered any output he disagreed with, and pointed to his "20-plus years of experience in the industry." The explosion killed three people and destroyed roughly 200 homes; the jury assigned the remaining 70 percent of the fault to Watson Grinding.
Moye said the reaction from other litigators has been immediate: "I've got lawyers all over the place saying… 'Holy shit, man. How did you get the prompts?'" Five of the analysts we track had circulated the story within a day, adding it to a running AI-ethics feed that has logged more than a hundred entries in the last three months. Knighthawk Engineering, Autenrieth, and 3M did not respond to 404 Media's requests for comment.
What others are reporting
-
Gadget Review Read →
Details how 350 pages of ChatGPT logs surfaced during a deposition recess; reports ChatGPT flagged its own '0% responsible' output as an easy target when used as mock opposing counsel.
They hired him and he used ChatGPT to write these reports, so really, ChatGPT was the expert in the case.
-
Blaze Media Read →
Adds KnightHawk Engineering's 'fake news' rebuttal and defense of AI as a legitimate professional tool; notes the ChatGPT logs were publicly searchable online before any subpoena was issued.
He acknowledged the prompts he put in were biased toward 3M to help 3M win the case [...] it's really egregious.
-
National Law Review Read →
Situates this as the third plaintiff-favorable verdict in four Watson Grinding trials; documents 3M's pre-explosion awareness of a malfunctioning gas-detection system spanning 2017 to 2019.
The gas-detection system was intended to provide a critical warning before a catastrophic release could endanger employees and the surrounding community.
Shared on Bluesky by 9 AI experts (top 5 by trust)
Originally reported by 404media.co
Read the original article →Original headline: 404 Media: 3M's Expert Witness Used ChatGPT for '85–90%' of Report in $61M Fatal Explosion Case