“Today, the Electronic Privacy Information Center (EPIC) released Generating Harms II: Generative AI’s New & Continued Impacts, a follow-up report that expands on the harms detailed in last year’s Generating Harms report. The new report includes a deep dive into four new issue areas and explores the different remedies and enforcement actions that are being considered by policymakers. Despite the constant attention in the past year and several proposals to address these harms, generative AI companies have not meaningfully contended with the real-life damage they are doing to individuals and society. Most discussions of generative AI focus on the technology’s outsized theoretical potential. Where harms are addressed, they are often hypothetical future harms that evoke science fiction. Unfortunately, harms stemming from GAI are not theoretical—they are widespread and happening right now. “The rapid spread of generative AI has created an explosion of disinformation, invasions of privacy, scams, and more,” said Calli Schroeder, lead of EPIC’s AI and Human Rights project. “This report details where those harms are expanding and identifies interventions necessary to rein in these systems. Ultimately, if generative AI cannot be deployed safely, then it should not be deployed at all.” The report details the scope of four new issue areas where generative AI is causing harms:

• Elections
• Privacy
• Data Degradation
• Content Licensing..”