All is not always as it appears – especially in the realm of electronically-stored information.  It has always been possible for ESI to be deliberately altered or entirely fabricated, but in recent years, new machine learning techniques have made it possible for computers to produce a tsunami of fake images, fake audio recordings, and fake video recordings with disturbing quality and ease.  First came “deepfakes” that enabled amateurs using personal computers to replace the original faces or voices in video or audio recordings with new ones, using publicly-available machine-learning algorithms trained using online source media.  Next came cloud-based generative AI tools, which allow for images and videos to be generated automatically from text prompts.  This looming tsunami of artificial content will likely pose real challenges to our current discovery processes, from new opportunities to inadvertently rely on something altered or unreal, to new opportunities for intentional spoliation or fabrication, to new technical and legal authentication challenges.  In this program, a group of expert practitioners will discuss these new challenges and how we can prepare for them.

• Learn the difference between deepfake tools, large language models, and other generative AI content sources
• Gain an understanding of the potential risks  of such materials in discovery and how you can work to mitigate them
• Consider how the existence of these technologies may undermine even legitimate video and audio evidence