OpenAI and Meta Speak Out: Deepfakes Will Severely Threaten the Reliability of Evidence in Legal Proceedings

According to DeepTech TechFlow, on October 20, Reuters reported that representatives from OpenAI and Meta stated during the U.S. Judicial Conference Advisory Committee on Evidence Rules meeting on the 15th of this month that AI-driven deepfakes will severely threaten the reliability of evidence in legal proceedings. Shira Anderson, Policy Advisor at OpenAI, said: "The quality of AI-generated images is continuously improving, and I believe that in the future, it will be impossible to identify them through scientific means." Devin Norder, Director of Product Management at Meta, added: "Humans can no longer distinguish between real media and media generated by at least the latest generations of cutting-edge AI models. Moreover, the cost and time required to create such media are indeed negligible." Experts warn that courts are increasingly struggling to guard against AI-generated fake evidence. Currently, several AI companies have added visible or invisible watermarks to AI-generated media, but there is no universal method for detecting these watermarks yet. However, the use of deepfakes in the judicial field remains rare: in a survey involving 931 U.S. judges, only 15 reported encountering situations where evidence credibility was questioned due to suspected deepfakes.