New Mexico lawyer sanctioned after filing ChatGPT-invented testimony
single source · 1 articles · safety · confidence: high · first seen 2026-09-11 19:34 UTC
A New Mexico defence lawyer has been sanctioned for citing testimony from witnesses who do not exist. The witnesses were generated by ChatGPT, and the lawyer submitted their statements in a court filing. When the fabrication came to light, the lawyer said he did not know that AI could invent facts, telling the court: "I didn't know that AI could hallucinate facts." Ars Technica reported the discipline on 11 September 2026. The article does not specify the exact penalty. The case is an example of a lawyer relying on a generative model's output without verifying it.
What this means for you
If you use a language model to draft legal filings or any public document, check every citation, name and quote yourself before it leaves your control. The model cannot tell you which parts it invented.
Key facts
- ·A defence lawyer in New Mexico was sanctioned for citing fake testimony in a court filing. source
- ·The testimony came from witnesses generated by ChatGPT and did not exist. source
- ·The lawyer said: "I didn't know that AI could hallucinate facts." source
- ·Ars Technica reported the case on 11 September 2026. source
What the sources say
- Ars Technica — Reports the sanction of a New Mexico defence lawyer who cited fabricated ChatGPT testimony, including the lawyer's explanation.
Sources
The original reporting. Follow these — they did the work.
- Ars Technica AIChatGPT-using lawyer punished for citing fake testimony from made-up witnesses2026-09-11