StreetInsider reported that OpenAI and Meta representatives told the US Judicial Conference’s Advisory Committee on Evidence Rules in Boston on 15 October that increasingly convincing deepfakes would make audio and video evidence harder to authenticate. OpenAI policy counsel Shira Anderson told the panel she did not believe a scientific method for detecting AI-generated images would ever exist.
The committee has examined deepfake authentication since 2023, alongside the reliability standards applied to AI-generated evidence. A Federal Judiciary Center survey conducted in March found that 15 of 931 responding judges had encountered a case in which a party challenged evidence as a deepfake. The panel’s chair, US District Judge Jesse Furman, said such challenges had not yet become common in court.
Meta’s director of product management, Devin Norder, told the committee that people could not distinguish media from recent AI models from authentic material, and that producing it took little time or money. California and the European Union have enacted watermarking requirements for deepfakes. Anderson said companies used different systems, leaving no universal way to detect every watermark, and that those systems would never be completely reliable.
