The reported use of ChatGPT by an expert retained by 3M shows why AI governance in high-stakes work is about more than polished writing: professionals must preserve independent judgment, verify every claim and source, and plan for records and disclosure. The episode is a case study, not a general court rule that AI use invalidates expert evidence.
How ChatGPT entered the 3M expert-witness case
The dispute arose from the January 2020 Watson Grinding explosion in Houston. 404 Media reported that a degraded, poorly crimped rubber welding hose leaked flammable gas before the explosion, citing the U.S. Chemical Safety and Hazard Investigation Board. The incident killed three people and damaged or destroyed homes, according to that reporting. Plaintiffs alleged that 3M’s work on a gas-detection system contributed to the explosion; that allegation should not be confused with an established finding.
As an Amazon Associate I earn from qualifying purchases.
404 Media reported that engineering expert Josh Autenrieth, associated with Knighthawk Engineering and retained by 3M, used ChatGPT to help prepare significant portions of an expert report. The outlet said discovery exposed prompts and related AI material, and that Autenrieth was questioned about the process. Among the prompt excerpts it reproduced were requests to “create an exceptional expert witness report defending the standard of care at 3M” and to “show how 3M is 0% at fault for the explosion at Watson Grinding.” These are reported prompt excerpts—not language from a court ruling. 404 Media’s account describes the reported prompts and discovery material.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesCBS News separately reported that the AI use drew scrutiny at trial and that legal experts raised concerns about reliability of testimony and evidence. That is reported commentary, not a judicial determination that the expert report was unreliable. CBS News’ report covers that scrutiny.
Why the prompts raise an independence question
An expert’s value depends on analysis grounded in evidence and professional judgment. A prompt asking a model to show that a client is “0% at fault” signals a desired conclusion before the work is complete. That creates an appearance problem even if a human later reviews and edits the output: readers may reasonably ask whether the report tested the evidence or organized it around an answer already chosen.
This is a governance inference from the reported prompt, not a finding by the court. AI use alone does not establish that an expert lacks independence. The important question is whether the expert can explain how the evidence, assumptions, methods, and reasoning support each conclusion—and whether that explanation is genuinely the expert’s own.
Rank #2
What the episode reveals about AI governance
Human accountability cannot be delegated to a model
An expert remains responsible for the report’s claims, reasoning, and conclusions, even when AI assists with drafting or analysis. A model’s fluent answer is not proof that an assertion is accurate, that an assumption is reasonable, or that the expert independently reached the conclusion.
Verification must include evidence and citations
Checking that prose reads well is not enough. Each factual statement, quotation, calculation, and citation needs to be traced to a source that actually supports it. The reporting’s reference to a “Citation Overlay” and discovery of AI interactions makes source traceability a practical concern: a neat citation format can conceal an unsupported claim if no reviewer checks the underlying material.
Rank #3
AI inputs and drafts may become litigation material
In this case, 404 Media reported that prompts and AI-related documents surfaced through discovery. That does not mean all prompts are automatically discoverable. Relevance, privilege, protective orders, and procedural context can affect what must be produced. The practical lesson is to treat AI records as potentially consequential and have counsel set appropriate retention and handling practices before use.
Rules work best before an expert begins
Organizations engaging experts can establish expectations in advance about permissible tools, confidentiality, records, review, and disclosure. These are risk-management recommendations drawn from the reported episode—not requirements shown to have been imposed by a court.
- Define whether AI tools may be used, and for which tasks.
- Protect confidential, privileged, or sensitive information by setting clear rules for what may be entered into a tool.
- Retain material prompts, outputs, and drafts in a way consistent with counsel’s instructions and applicable obligations.
- Require the expert to verify sources, quotations, calculations, and factual claims independently.
- Agree who reviews AI-assisted material and what method of use should be disclosed.
A practical review framework for AI-assisted expert work
Before relying on an AI-assisted report, the expert and retaining organization can use these questions to test whether the process supports accountable analysis:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Independence: Did the expert develop conclusions from the evidence, rather than ask a model to justify a predetermined result?
- Traceability: Can each material claim be connected to a source, and does that source support the claim as written?
- Responsibility: Can the expert explain the assumptions, methods, and reasoning without relying on the model’s authority?
- Records: Are prompts, outputs, and relevant drafts handled under a retention plan set with counsel?
- Disclosure: Have the client and expert agreed in advance how material AI assistance will be reviewed and described?
- Confidentiality: Were inputs handled in a way consistent with legal, contractual, and case-specific obligations?
What the case does—and does not—establish
The reporting describes one expert’s use of ChatGPT in one high-stakes dispute, with related material reportedly emerging in discovery and drawing questions at trial. It supports a practical warning: AI-assisted work can complicate confidence in expert independence when prompts appear to seek a client-favorable conclusion, and weak source checks can make polished text difficult to trust.
It does not establish a universal prohibition on AI-assisted expert work, a general rule that prompts are always discoverable, or a court holding that AI use makes testimony unreliable. The available reporting also does not establish the final judgment or appeal status; claims about the case’s ultimate legal outcome require confirmation from the relevant Harris County court record.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
