Extract from Wortzman and Rothman’s article “Maduro Photos Show the Challenge of Determining Whether Evidence is AI-Generated”
Deep fakes (highly realistic fake images and videos generated by artificial intelligence) are increasingly complicating the work of legal professionals and the Courts. When the New York Times received three photos of a captured Nicolas Maduro, it had to quickly determine if any of the photos were AI generated before it went to print. Thanks to the NYT for the detailed reporting on the process and challenges they faced. As deep fakes become increasingly sophisticated and accessible, distinguishing between genuine and manipulated evidence is more challenging—and necessary—than ever. Lawyers are facing this head on as digital evidence is exchanged in litigation and investigative matters in the age of AI.
The NYT process highlighted the complexities involved in authenticating images that may have been manipulated or entirely fabricated by AI. In the short time before going to print, the NYT could not verify that any of three Maduro photos they had received were authentic, underscoring that the process is painstakingly intricate and requires expertise and technology to get to the right answer. This scrutiny is vital in the legal context, as the reliance on deep fakes can have far-reaching implications for administration of justice.
