The Department of Justice Is Using AI Detector Pangram Against Defense Attorneys
Reported by 404 Media and aiweekly2 sources ↓
In a Texas voter-fraud case, DOJ lawyers submitted Pangram results to argue that defense filings relied heavily on AI. The defense says it used AI under time pressure after a lawyer’s medical emergency.
Key points
- Prosecutors filed Pangram results as exhibits and argued they should not have to respond substantively to the briefs
- The detector results were mixed and did not indicate the briefs were entirely AI-drafted
- Judge Mark Pittman criticized the defense filings as arguments clearly generated by AI
- Defense attorney Mario Herrera says a colleague’s medical emergency left him four days to prepare for trial
Sources
Details
The case involves Khalwinder Bhengura, who was convicted of illegally registering to vote and voting in the 2024 election after falsely stating he was a U.S. citizen on a driver’s license application. The Northern District of Texas requires attorneys to disclose generative AI use in filings, and Bhengura’s attorneys disclosed using it.
DOJ lawyers said Pangram’s results supported their claim that little substantive content in the defense filings had been written by a human. Herrera told 404 Media that a medical emergency involving lead attorney Nicole Knox left him to prepare for a federal jury trial in four days; he said he used available legal tools and did not violate rules or laws.
Written from coverage as of
Images
- 404 Media
- 404 Media
- 404 Media