On September 9, 2026, Paul Christiano joined the OpenAI Safety and Security Committee. He estimates a 4% risk of catastrophic loss of control over AI within one year and a 15% risk within three years. He argues that AI systems capable of automating their own development could create a feedback loop, leading to rapid, potentially uncontrollable progress.
Anthropic CEO Dario Amodei published an essay proposing a three-part plan to pace AI development. He suggests allowing independent external auditors access to internal processes and safety data, with the right to publish findings. Amodei also calls for coordinated industry safety standards and government-backed legal frameworks to allow companies to discuss safety measures without antitrust concerns.
OpenAI CEO Sam Altman expressed support for the proposal, noting that OpenAI plans to implement similar external audits. Elon Musk also publicly endorsed Amodei's position. Reuters reported that both Altman and Musk support these regulatory efforts, though the industry remains divided on the necessity of new government-imposed restrictions.
FTC Commissioner Andrew Ferguson warned on September 15 that industry requests for new regulations and antitrust exemptions could create barriers to entry, protecting current market leaders. He noted that such rules might impose high fixed costs that only established companies can afford, potentially stifling competition from startups.
Donald Trump publicly rejected the idea of new AI regulations, arguing that existing laws are sufficient. He criticized calls for slowing down development, linking them to the technological race with China. Trump stated that the only necessary control for AI is strong presidential leadership, opposing the industry's push for new state-mandated guardrails.
