The brief consensus on AI safety among tech leaders has fractured. Just days after Anthropic CEO Dario Amodei proposed a three-step plan for slowing development, including third-party evaluators and international agreements, the industry is split. OpenAI’s Sam Altman and Google DeepMind’s Demis Hassabis initially signaled support. But that unity collapsed quickly. Meta’s Mark Zuckerberg, Nvidia’s Jensen Huang, and SpaceX’s Elon Musk actively scuttled proposals for an industry-funded independent regulator. The Wall Street Journal reported this move killed the idea of an AI equivalent to FINRA, the financial sector’s private watchdog.
Political headwinds are making federal oversight unlikely. The Trump administration has dismissed AI safety concerns as a “hoax.” President Trump recently called Huang during a conference to declare that robots will not take over. He posted that the only guardrails needed are a “STRONG AND SMART (High IQ!) PRESIDENT.” This stance contradicts his own first-term policy. Executive Order 13960, signed by Trump in 2020, established principles for federal AI use to protect privacy and civil liberties. Nick Reese, a former DHS policy director, noted the stark shift. “Six years ago, the Trump Administration signed executive action on AI safety,” Reese said. “There is a real change in the president’s position on the topic.”
OpenAI remains persistent. Chris Lehane, the company’s global affairs chief, told The Verge that government has an important role. “The AI policy window is open, with growing bipartisan support for mandatory national safety standards,” Lehane said. He urged Congress to turn momentum into a framework that keeps the U.S. leading in AI. An anonymous venture capitalist compared the current moment to the early days of cars and airplanes. Those industries were once unsafe and unruly. They only achieved widespread adoption after creating collective safety standards. The analogy suggests regulation might eventually help, not hinder, growth.
For now, however, autonomy wins. Zuckerberg distanced himself from slowdown calls on X, stating every lab has the responsibility to move at its own pace. Anthropic declined to comment, though Amodei has long advocated for well-considered regulation despite accusations of “doomerism.” With major players like Safe Superintelligence Inc. and Thinking Machines Lab staying silent, the path forward is unclear. The industry is betting that self-policing will suffice, even as the political environment grows increasingly hostile to external oversight.
