Key facts
- Major AI labs are pursuing state-level legislation to establish AI safety standards as federal action stalls.
- OpenAI and Anthropic are endorsing and influencing state bills, aiming to create a de facto national regulatory framework.
- Illinois passed SB 315, mandating safety assessments and third-party audits for large AI developers.
- California and New York have enacted similar AI safety laws, with OpenAI engaging with lawmakers on these bills.
- Smaller tech firms express concern that varying state regulations could lead to a complex compliance environment.
Major artificial intelligence firms are taking a proactive approach to regulation by championing state-level legislation, aiming to establish national AI safety standards in the absence of federal action. OpenAI and Anthropic, two leading AI labs, are strategically endorsing and influencing bills in states like Illinois, California, and New York to create a de facto national framework.
OpenAI's chief of global affairs, Chris Lehane, described the strategy as a form of "reverse federalism," where states replicate each other's AI safety measures, potentially paving the way for federal legislation. The company has seen success with bills like Illinois's SB 315, which mandates safety assessments and third-party audits for large AI developers with over $500 million in revenue. Similar legislation has passed in California and New York, with OpenAI expressing support for their outcomes.
Anthropic, while also supporting federal regulation, has adopted a similar strategy, endorsing bills like New York's RAISE Act and Illinois's SB 315. The company views these state endorsements as a crucial first step in the broader national conversation about AI regulation. However, smaller and mid-sized tech firms are concerned that this state-by-state approach could lead to a complex and fragmented regulatory landscape without a clear federal preemption.
Despite efforts by the Trump administration to push for federal preemption, lawmakers in Congress are still navigating proposals, indicating a lengthy process ahead. The AI firms' strategy acknowledges the slow pace of federal action and the states' eagerness to address growing public concerns about AI's impact on jobs, security, and the environment.