Anthropic’s safety warnings may have just backfired — the government has pulled the plug on its

Anthropic shuts down its most powerful AI models in the US. The company complied with a government directive to immediately disable access to Claude Fable 5 and Claude Mythos 5, citing national security concerns.

The US government's decision to order Anthropic to shut down its most powerful AI models has sent shockwaves throughout the tech industry. On Friday, the company announced that it had received a directive to disable Claude Fable 5 and Claude Mythos 5, two of its most advanced AI models, due to national security concerns. Anthropic has complied with the order, but the company has made it clear that it disagrees with the government's decision.

Mythos, Anthropic's most capable AI model, was first previewed in early April and has been kept tightly restricted due to its exceptional ability to find security vulnerabilities in software. The company had launched a controlled program called Project Glasswing, sharing Mythos with roughly 50 vetted organizations, including Amazon, Apple, Google, Microsoft, and CrowdStrike, to use for defensive cybersecurity work. Fable 5, on the other hand, was released just three days ago and was touted as a version of Mythos fitted with guardrails that block responses in high-risk areas like cybersecurity and biology, making it safe enough for general release.

The government's directive is framed as an export control action, restricting foreign national access to the models. However, Anthropic claims that the underlying concern is a claimed jailbreak of Fable 5. The company says that the government has provided only verbal evidence of a "potential narrow, non-universal jailbreak" – one that amounts to prompting the model to read a specific codebase and identify software flaws. Anthropic argues that this level of capability is already widely available in other publicly accessible models, including OpenAI's GPT-5.5, and is used routinely by cybersecurity professionals for defensive purposes.

Anthropic's broader argument is that its strongest safeguards operate through independent classifier systems that function separately from the model itself, meaning that even if someone convinces Fable to keep talking past a refusal, the underlying protections against the most dangerous outputs remain in place. The company believes that the government's decision to shut down its models is unwarranted and could have far-reaching implications for the development of AI technology.

The shutdown of Anthropic's AI models has significant implications for the company's business and its plans for an initial public offering (IPO). Anthropic has staked much of its public identity on being the safety-conscious alternative to its rivals, and the government's intervention has raised questions about the company's ability to balance safety concerns with the need to innovate and stay competitive. The incident has also sparked a debate about the role of government regulation in the development of AI technology and the need for clearer guidelines and standards for the industry.

In a lengthy blog post, Anthropic expressed its frustration with the government's decision and argued that the shutdown of its models could have a chilling effect on innovation in the AI industry. The company believes that the standard applied by the government in this case could essentially halt all new model deployments for all frontier model providers. Anthropic's concerns are shared by many in the industry, who argue that overregulation could stifle innovation and hinder the development of AI technology.

The incident has also sparked a war of words between Anthropic and its rival, OpenAI. OpenAI's CEO, Sam Altman, has been critical of Anthropic's handling of Mythos, which he has described as "fear-based marketing." Altman has argued that Anthropic's decision to restrict access to Mythos was a marketing ploy designed to generate buzz and attract attention to the company's products. Anthropic, on the other hand, has defended its decision to restrict access to Mythos, arguing that the model's exceptional capabilities made it too dangerous to release to the public.

The shutdown of Anthropic's AI models has significant implications for the future of AI development and the role of government regulation in the industry. As AI technology continues to evolve and become more powerful, the need for clear guidelines and standards for the industry will become increasingly important. The incident has highlighted the need for a more nuanced and informed approach to regulating AI technology, one that balances the need for safety and security with the need to innovate and stay competitive.

In the coming weeks and months, Anthropic and other AI companies will be watching closely to see how the government's decision to shut down its models will impact the industry. The company has expressed its intention to pursue an IPO this year, and the shutdown of its models could have significant implications for its valuation and attractiveness to investors. As the AI industry continues to evolve and grow, the need for clear guidelines and standards will become increasingly important, and the shutdown of Anthropic's AI models has highlighted the need for a more nuanced and informed approach to regulating AI technology.