OpenAI, Meta caution US judiciary about risk of 'deepfake' evidence

Live updates

1. 1h ago

   OpenAI and Meta Warn US Judiciary About AI Deepfake Evidence

   Representatives from OpenAI and Meta Platforms warned a federal judicial panel in Boston that artificial intelligence-generated deepfakes have advanced to a level where verifying the authenticity of audio and video evidence in court will grow increasingly difficult. Speaking to the US evidence rules committee as judges consider proposed changes to evidence rules, tech company representatives cautioned that current detection methods may soon become ineffective. While actual cases involving manipulated digital evidence remain rare so far, the growing sophistication of synthetic media poses a significant challenge for courts and police preparing to combat fake evidence.

   Why it matters

   The warning highlights growing vulnerabilities in the legal system as artificial intelligence tools become more accessible and convincing. Judges are currently evaluating proposed changes to evidence rules, specifically Rule 901(c), to address the rising threat of manipulated media. Legal experts and police agencies face mounting pressure to adapt as fake images, audio, and video threaten the reliability of courtroom exhibits.

   What is confirmed

   • Representatives from OpenAI and Meta Platforms told a federal judicial panel on Thursday that AI-generated deepfakes have become so advanced that verifying the authenticity of audio and video evidence in court will grow more difficult.
   • The warning was delivered to a U.S. judicial panel in Boston as judges consider proposed changes to evidence rules.

   Still unconfirmed

   • Actual cases involving deepfake evidence in courtrooms are rare so far.

   What to watch next

   • Decisions by the US evidence rules committee regarding proposed changes to Rule 901(c).
   • Further developments in digital forensics and detection methods for AI-generated deepfakes in law enforcement.
   Sources used for this update (10)

   1. Reuters — OpenAI, Meta caution US judiciary about risk of 'deepfake' evidence
   2. Orange County Register — Fake AI-generated images are more convincing than ever. Are courts and police ready to combat them?
   3. Magzter — Objection, Your Honor. The Witness Isn't Real | Newsweek US - news
   4. Crowell & Moring LLP — “Evidence in the Age of AI: Digital Evidence, Deepfakes and Manipulated Evidence,” October 10, 2026, Association of Business Trial Lawyers 52nd Annual Seminar
   5. Durham Post — AI, deepfakes and divorce in the age of fake evidence
   6. www.ibtimes.co.uk — OpenAI Warns Courts May Have No Reliable Way to Spot AI Deepfake Evidence | IBTimes UK
   7. www.yahoo.com — OpenAI and Meta warn U.S. judicial panel about deepfake evidence
   8. oecd.ai — OpenAI and Meta Warn U.S. Judiciary of Deepfake Evidence ...
   9. uk.finance.yahoo.com — OpenAI and Meta warn U.S. judicial panel about deepfake evidence
   10. www.explainx.ai — OpenAI, Meta Warn Judges on Deepfake Evidence (Rule 901c ...
   confidence 100%