Cybersecurity Experts Sign Open Letter Demanding U.S. Government Lift Mythos AI Restrictions
More than 120 cybersecurity professionals, technologists, and artificial intelligence researchers have signed an open letter calling on the United States federal government to lift export control restrictions on Anthropic's Mythos and Fable large language models. The letter also urges officials to develop a more transparent and collaborative process for evaluating AI-related national security risks going forward.
The move reflects growing tension between the government's desire to control potentially powerful AI technologies and the private sector's need to develop and deploy them freely — a debate that is rapidly shaping the future of AI policy in the United States and abroad.
What Triggered the Open Letter?
The open letter, dated June 14, was addressed directly to Commerce Secretary Howard W. Lutnick and National Cyber Director Sean Cairncross. It followed a significant announcement made the previous Friday, June 13, in which Anthropic disclosed that it had disabled certain access to its Fable 5 and Mythos 5 AI models in response to a federal export control directive.
That directive instructed Anthropic to suspend model access for "any foreign national," a sweeping definition that applied not only to users outside the United States but also to foreign national employees working within the company. The breadth of the restriction alarmed many in the technology and cybersecurity communities, who viewed it as overly broad, potentially counterproductive, and damaging to America's competitive standing in global AI development.
Anthropic had launched Mythos 5 and Fable 5 just four days before the suspension, on June 9. At the time of launch, the company emphasized that it had built robust internal safeguards specifically designed to prevent misuse of the models in sensitive domains including cybersecurity, biology, chemistry, and AI model distillation. Despite those precautions, the government moved quickly to issue the export control directive, prompting the community backlash that culminated in the open letter.
Who Is Behind the Open Letter?
Among those who publicly amplified the letter was Joe Levy, CEO of the well-known cybersecurity firm Sophos, who shared a link to the open letter and voiced his support for lifting the restrictions. His involvement signals that this is not simply an academic or advocacy-driven effort — it has real buy-in from operational leaders in the cybersecurity industry who work on the front lines of defending organizations from digital threats.
The open letter itself is hosted at freefable.org, and has gathered over 120 signatories spanning cybersecurity professionals, AI researchers, software engineers, and other technology experts. The signatories argue that restricting access to advanced AI models like Mythos does not meaningfully reduce national security risks, and may in fact harm the United States by slowing the development of defensive cybersecurity tools that rely on cutting-edge AI capabilities.
The Broader Debate: AI Safety vs. AI Access
The controversy surrounding Mythos and Fable is part of a much larger and increasingly urgent debate about how governments should regulate powerful AI systems. On one side, national security agencies and policymakers worry that frontier AI models could be exploited by foreign adversaries to develop cyberweapons, accelerate bioweapons research, or otherwise destabilize critical systems. These concerns are not unfounded — the rapid advancement of large language models has created genuinely novel risks that policymakers are still struggling to define and address.
On the other side, AI developers and cybersecurity professionals argue that overly aggressive restrictions stifle innovation, put American companies at a competitive disadvantage against foreign rivals who face no such limitations, and ultimately make the country less secure by denying domestic defenders access to the best tools available.
The signatories of the open letter appear to fall firmly in the latter camp. They contend that the export control directive issued against Anthropic was applied too bluntly, without adequate consultation with the industry, and without a clear framework for evaluating when such restrictions are truly necessary and proportionate.
What the Open Letter Demands
Beyond simply asking for the restrictions to be lifted, the letter calls on the federal government to commit to a new, more rigorous and inclusive process for assessing AI risk. Specifically, the signatories want officials to engage meaningfully with the cybersecurity and AI research communities before issuing directives that could have sweeping consequences for the industry.
This demand for procedural reform may ultimately prove more significant than the immediate question of Mythos and Fable access. If adopted, a more collaborative risk-assessment framework could help the government make better-informed decisions about AI regulation while giving the private sector a legitimate avenue to provide expertise and push back against overbroad restrictions.
Anthropic's Position and the Road Ahead
Anthropic finds itself in a difficult position — a company that has publicly staked its identity on responsible AI development, and that launched Mythos with explicit safety features in place, now caught between federal directives and a growing chorus of industry voices demanding unrestricted access to its flagship models.
How the Commerce Department and the National Cyber Director respond to the open letter will be closely watched across the AI and cybersecurity sectors. The outcome could set important precedents for how the U.S. government handles the regulation of future advanced AI systems.
For now, the restrictions remain in place, and the debate over where to draw the line between AI safety and AI freedom continues to intensify.
Key Takeaways
- More than 120 cybersecurity and technology professionals have signed an open letter asking the U.S. government to lift export controls on Anthropic's Mythos and Fable AI models.
- The directive required Anthropic to suspend access for all foreign nationals, including its own employees, just days after the models launched.
- Sophos CEO Joe Levy is among those publicly supporting the letter, highlighting significant industry opposition to the restrictions.
- The open letter also calls for a new, more transparent government process for assessing AI risk in consultation with the private sector.
- The controversy reflects a deepening national debate about how to balance AI safety with the need for open access to powerful AI tools.