California Leads AI Regulation with SB 1047 and AB 2013: What AI Developers Need to Know

Source: WilmerHale

By Priya Khosla September 19, 2024

Artificial intelligence (AI) regulation is gaining momentum in U.S. state legislatures, with California taking the lead by passing two significant AI bills—SB 1047 and AB 2013. SB 1047, known as the "Safe and Secure Innovation for Frontier Artificial Intelligence Models Act," mandates that developers of large AI models implement comprehensive safety measures to mitigate potential risks. Meanwhile, AB 2013 requires transparency regarding the training data used in generative AI systems. Both bills are currently awaiting approval from Governor Gavin Newsom, reflecting California’s proactive stance on regulating AI technology.

SB 1047 introduces strict safety protocols for developers who train advanced AI models in California, requiring them to follow a duty of reasonable care to prevent harm, including cybersecurity protections and detailed safety procedures. The bill also includes provisions for third-party audits and whistleblower protections, highlighting the state’s concern for ethical AI development. AB 2013, on the other hand, focuses on transparency, requiring developers to disclose details about the data used to train AI systems, including the sources and legality of the data.

If signed into law, these bills could have widespread implications, not only for companies operating in California but for AI developers across the country. As more states begin to propose similar laws, and with international regulatory efforts like those in the European Union gaining traction, companies must adapt to a rapidly evolving regulatory landscape. Even if these specific California bills don’t pass, their provisions offer a glimpse into the future of AI regulation, emphasizing safety, transparency, and accountability.