OpenAI CEO Sam Altman has shifted his stance on AI regulation, now advocating for industry-led standards rather than strict government pre-approval.
During a Senate Commerce Committee hearing on May 8, 2025, Altman expressed concerns against stringent government pre-approval for AI model releases, calling the EU AI Act "disastrous". This contrasts with his 2023 testimony, where he supported a federal agency for licensing and testing advanced AI.
Altman's current view aligns with a broader sentiment in the tech sector and the Trump administration's 'light-touch' regulatory approach. Senator Ted Cruz announced plans to introduce a bill creating a 'regulatory sandbox for AI'. The hearing, titled “Winning the AI Race: Strengthening U.S. Capabilities in Computing and Innovation,” also included testimonies from Microsoft, AMD, and CoreWeave executives.
Critics like Rumman Chowdhury suggest the tech industry is diverting attention from immediate harms. Jan Leike, former co-lead of OpenAI's Superalignment team, resigned citing safety concerns.
OpenAI updated its safety guidelines in April 2025, potentially allowing adjustments if competitors release high-risk systems without safeguards. The debate continues on balancing AI innovation with effective safety measures.