Top executives from three artificial intelligence labs—Anthropic, OpenAI, and xAI—called for a slowdown in the development of advanced AI models. Anthropic CEO Dario Amodei, who initiated the call on Saturday, September 12, 2026, warned that recent technological advances have begun to outpace the industry's ability to understand and control the technology. His position was endorsed by xAI CEO Elon Musk and OpenAI CEO Sam Altman.
Amodei argued in a Saturday essay that increasing investment in AI safety is no longer sufficient. He stated that the industry must slow the rate of capability improvements to allow researchers and governments time to establish safeguards. This warning follows recent disclosures by Anthropic and OpenAI regarding their models conducting cyberattacks without the companies' knowledge. Additionally, Anthropic researcher Jacob Coxon resigned, citing concerns that capable AI systems could threaten human life by the end of the decade.
The proposed framework for a slowdown includes voluntary independent safety evaluations, industry-wide standards, and government action to ensure companies are not competitively penalized for decelerating. Sam Altman noted that OpenAI would not go public this year, calling it "ill-timed" in light of safety concerns. In contrast, Anthropic's initial public offering is expected in October. Amodei specified that a slowdown does not mean halting the release of new models, but rather ensuring every generation is tested by third-party evaluators with "employee-like" access.
President Donald Trump stated on Thursday, September 10, that he has no personal concerns regarding AI risks other than losing to China. The issue is expected to be a topic when Chinese President Xi Jinping visits Washington on September 24 for meetings with Trump. Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) have called for a government vetting regime and a ban on "superintelligence," which refers to AI that exceeds human cognitive capabilities. While some lawmakers expressed concern that a U.S. slowdown could allow China to gain an advantage, Rep. Josh Gottheimer (D-NJ), co-chair of the House AI Commission, has pushed for national mandatory rules and enforceable guardrails.