California is at the forefront of a heated debate over artificial intelligence regulation. AI startups and tech giants fiercely oppose a bill, SB 1047, progressing through the state legislature. They argue that the proposed regulations are overly vague and could stifle innovation. While some in the industry advocate for federal oversight, they call for more precise guidelines than those in the California bill.
Key Provisions Of SB 1047
SB 1047, which has already passed the state Senate and two Assembly committees, mandates that developers of large AI models—those costing over $100 million to train and meeting a specified computing power threshold—must conduct safety tests to prevent "catastrophic harm."
This includes damage exceeding $500 million or mass casualties. Additionally, AI systems must have a fail-safe mechanism for human intervention if they behave dangerously. This bill impacts any company operating in California, regardless of its headquarters.
Industry Reactions And Concerns
Opponents, including major tech companies like Meta and OpenAI, argue that the bill’s broad and ambiguous language could hamper innovation. They worry about the lack of detailed guidance on safety tests and the composition of the oversight commission. Microsoft, for example, criticized the bill for potentially leading to misguided investments and ineffective regulations.
