Anthropic CEO Dario Amodei warned Mythos posed a national security threat. Washington just responded.

Dario Amodei, the CEO of Anthropic, has been vocal about the potential dangers posed by artificial intelligence, particularly his company's latest model, Mythos. In a recent essay, Amodei expressed that AI's capabilities are now 'undeniable' and highlighted significant risks associated with cybersecurity, the financial sector, and national security. He called for stronger government regulation to mitigate these threats. In a dramatic turn of events, the U.S. government responded to his warnings by intervening and ordering Anthropic to restrict access to Mythos 5 and Fable 5 for foreign entities. This decision came as a surprise and stirred a considerable reaction within the tech community. The Pentagon's chief information officer supported the action, emphasizing that some priorities outweigh profit-driven motives. The tech industry reacted swiftly, with AI researcher Gary Marcus criticizing the government's response as excessively alarmist and counterproductive. Meanwhile, Yann LeCun, a prominent figure in AI development, pointed fingers at Amodei for instigating fear around his own products, suggesting that his warnings have now led to unintended consequences. Amodei, who previously worked at OpenAI before founding Anthropic, has positioned himself as a responsible figure in the AI landscape, raising alarms about the potential for AI to disrupt job markets and societal norms. He has warned that AI could displace a significant number of entry-level jobs, drawing parallels with the economic impacts of past crises. Despite recent efforts to soften his stance on employment impacts as Anthropic prepares for a public offering, Amodei continues to highlight serious safety concerns. He cautioned that the risks associated with Mythos are just the beginning, predicting potential biological risks and serious autonomy issues in the near future. Earlier this month, Anthropic advocated for a temporary halt on the development of advanced AI models, arguing that current advancements could lead to humans losing control over AI systems. This call for a pause was answered on Friday with the government's intervention, at least for Anthropic's projects.