California Enacts AI Transparency Act, Mandates Training Data Disclosure to Combat Deceptive AI Content

Sep 29, 2024, 07:13 AM

California Governor Gavin Newsom has signed AB 2013, known as the California AI Transparency Act, on Saturday. This law requires generative AI companies providing services to Californians to publicly disclose summaries of the data used to train their AI models. The new legislation is considered the most comprehensive set of rules in the United States for disclosing data fed into AI models and aims to combat AI-generated deceptive election content. Some observers note that the act echoes the 'Brussels Effect,' potentially setting global standards for AI regulation similar to how the EU's GDPR influenced privacy laws worldwide. Chris Manning, a prominent figure in the AI community, commented, "There is huge uncertainty about generative AI. Legislation that brings sunlight is great; premature restrictions of dubious soundness are not." The law is seen as a step toward greater transparency amid growing uncertainty about generative AI, with proponents arguing that such legislation can drive responsible AI development globally.