Anthropic has disabled access to its most advanced AI systems following a US government export control directive requiring the suspension of all use of Mythos 5 and Fable 5 by foreign nationals, including employees and users located inside and outside the United States.
“The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance,” Anthropic said in its official statement dated June 12, 2026.
The company confirmed that the directive was received at 5:21 pm (ET) and that access to all other Anthropic models would remain unaffected.
“The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees,” the statement said.
Government Cites National Security and Alleged Jailbreak Risk
US authorities did not provide detailed public justification for the order. Anthropic said its understanding is that the directive followed concerns that the government had identified a potential method of “jailbreaking” Fable 5.
“Our understanding is that the government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5,” the company said.
Anthropic said it reviewed a demonstration of the technique and found it revealed “a small number of previously known, minor vulnerabilities.”
“These vulnerabilities all appear relatively simple, and we have found that other publicly-available models are able to discover them as well without requiring a bypass.”
Anthropic: Safeguards, Testing and Red Teaming Efforts
In its statement, Anthropic detailed extensive pre-release testing and safety work conducted in collaboration with government and third-party organizations.
The company said: “We have instituted strong safeguards that greatly reduce the likelihood that Fable is misused for tasks related to cybersecurity (among others). In fact, our safeguards are so strong that many users have complained that they are overly broad.”
It added:
“In the weeks leading up to the launch of Fable, Anthropic worked with the US government, the UK AISI, multiple private third-party organizations and internal teams to red-team Fable’s safeguards for thousands of hours in total.”
According to Anthropic, those tests showed:
“These tests showed that Fable’s safeguards are substantially more effective than those of any previously deployed model.”
The company also stated: “No testers have yet been able to find a universal jailbreak—a jailbreak method that can very broadly bypass the model’s safeguards, unblocking a wide range of cyber capabilities.”
Defense-in-Depth Strategy and Industry Risk Argument
Anthropic emphasized that it does not believe perfect jailbreak resistance is achievable and said its mitigation strategy relies on layered protections.
“We suspect that perfect jailbreak resistance is not currently possible for any model provider. Every safeguard used in the industry is vulnerable to non-universal jailbreaks,” the company said.
It added: “Given that perfect jailbreak resistance does not appear to be possible today, Anthropic adopted a defense in depth strategy with Fable 5.”
The company said the approach was designed so that jailbreaks would either be narrow or difficult to scale, supported by monitoring systems intended to detect misuse.
“This is also why Anthropic has required 30-day retention of customer data with Fable—a policy change that carries real costs for us with customers, but that allows us to research and mitigate jailbreaks.”
Anthropic further stated: “We stand by this defense in depth strategy. It reduces the risks posed by Fable, making them comparable to the risks of existing models already deployed across the industry.”
Dispute Over Government Assessment
Anthropic said it had not received evidence of any significant harmful exploit linked to the alleged vulnerability. “We have not even received a disclosure of a concerning non-universal potential jailbreak that led to a harmful result.”
The company added: “To date, the government has only given us verbal evidence of a potential narrow, non-universal jailbreak.”
Anthropic said it reviewed the underlying report and concluded: “We have reviewed a report that we believe is the basis of the government’s directive and validated that the level of capability displayed there is widely available from other models (including OpenAI’s GPT-5.5), and is used every day by the defenders who keep systems safe.”
The company also signaled it would provide further technical details within 24 hours.
Compliance, Objections, and Industry Impact Warning
While complying with the directive, Anthropic disputed the decision to revoke access to commercially deployed systems.
“We are complying with the government’s legal directive and are removing access to Fable 5 and Mythos 5 for all users,” the company said.
However, it added: “We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.”
Anthropic warned that applying such a standard broadly could halt industry progress: “If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”
The company also stated that it believes the decision does not follow principles of transparency and technical grounding: “This action does not adhere to those principles.”
Anthropic concluded its statement with an apology to users and an indication that it is seeking restoration of access: “We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.”
Internal and Global Access Restrictions Intensify Fallout
The directive extends to foreign nationals globally, including Anthropic employees, effectively removing access for a significant portion of researchers and developers working with the systems.
The restriction has triggered wider debate over AI governance, export controls, and whether advanced models should be treated as strategic technologies subject to national security restrictions similar to dual-use defense systems.
Europe Reacts to AI Access Restrictions
European policymakers and political figures described the decision as a defining moment for digital sovereignty, arguing that reliance on US-controlled AI systems exposes governments and industries to external policy shifts.
Bruno Retailleau, former French interior minister, said the move should serve as a “wake-up call,” adding that dependence on foreign technology leaves nations vulnerable to being “unplugged overnight.”
Benjamin Haddad, French minister delegate for Europe, called the decision “an accelerator of the geopolitical battle over AI,” urging Europe to expand investment in domestic innovation.
Former French prime minister Édouard Philippe said AI has become “a critical infrastructure, as essential as electricity or the Internet.”
Industry and Policy Backlash
Across policy and industry circles, the directive has intensified debate over export controls, AI governance, and the classification of frontier models.
Dean W. Ball, senior fellow at the Foundation for American Innovation, said: “If this is true, it is just baffling. An administration whose posture is that we should export advanced AI chips to China, which also wants to ban… Britain (and every other non-American on Earth)… from using our best models? I have no words.”
Chris McGuire, senior fellow for China and emerging technologies at the Council on Foreign Relations, said: “I actually think targeted export controls on model access are prudent. But across the board controls on all countries on a single model, without any warning, is highly questionable.”
He added: “If BIS doesn’t understand how to use its authorities or what the implications are of its actions, then it needs to find some new personnel who can actually execute a competent export control strategy.”
Matthew Pines, CEO of Physical Superintelligence, wrote: “this is gonna send shockwaves through every lab and neolab… U.S. export control laws operate under a strict liability standard… they are a very sharp blade…”
Dan Shipper, CEO of Every, said: “my take on this situation currently is that they’ll unban it in a few days and the net effect will be increased demand for Fable”
He added: “however this kind of thing is extremely disruptive and distracting for people inside of the company.”