Blab Tech
US Export Control Order Forces Anthropic to Suspend Global Access to Fable 5 and Mythos 5 Models, Prompting EU Tech-Sovereignty Push
On June 12, 2026, a U.S. export‑control directive halted Anthropic’s flagship language models worldwide. The Commerce Department’s order required the company to disable its Fable 5 and Mythos 5 models for every foreign user, regardless of location. This marked the first publicly confirmed instance in which a U.S. government order forced a frontier AI provider to remove commercial model access on a global scale.
Anthropic, an American AI startup founded in 2021, launched Fable 5 and Mythos 5 on the very day the directive arrived. In a statement released that afternoon, the company explained that the letter “did not provide specific details of the national security concern” and that compliance meant disabling both models for all customers. Anthropic also noted that U.S. officials had become aware of a method for bypassing, or “jailbreaking,” Fable 5. The firm reviewed the technique, disagreed with the finding, and warned that applying such a standard across the industry would “essentially halt all new model deployments for all frontier model providers.”
The ban sent shockwaves through the European Union, where researchers and businesses had already integrated the models into academic and commercial workflows. Earlier in June, Anthropic had granted the European Union’s cybersecurity agency ENISA access to Mythos under the Glasswing project—a partnership now at odds with the sudden suspension.
In response, European Commission Executive Vice‑President for Tech Sovereignty, Security and Democracy Henna Virkkunen posted on social media on June 15 that “Europe represents an economic opportunity, not a security risk. We are and will remain a trusted partner. This is why we need to cooperate on emerging, powerful AI models. This is a shared global challenge.” The comment followed the Commission’s announcement that it was assessing the practical impact of the directive on EU users and cautioned that measures to address risks from advanced AI should not be discriminatory against partners.
Virkkunen said the episode “underlines the need for Europe’s technological sovereignty” and urged swift adoption of a package of measures announced earlier in June. The package includes procurement preferences for European firms in sensitive public‑cloud and AI contracts. It will become law only after negotiation with the European Parliament and member states. The European Union currently sources more than 80 percent of its digital products, services, infrastructure and intellectual property from outside the bloc—a fact that has amplified concerns about digital dependence.
The directive raises several questions for stakeholders. Observers note that the U.S. government has not yet provided technical details backing the order, and it remains unclear whether Anthropic will restore access once the directive is lifted. The European Parliament’s speed in moving on the proposed procurement measures will also influence EU procurement strategy. Additionally, it remains to be seen whether other frontier model providers will face comparable export‑control orders.
The incident underscores the growing regulatory friction between U.S. export controls and the global AI ecosystem. It also highlights the EU’s push for greater technological sovereignty, a policy direction that could reshape how European institutions and enterprises manage vendor risk and compliance across jurisdictions.
At present, Anthropic has not announced a timeline for reinstating Fable 5 or Mythos 5. The European Commission continues to monitor the situation, and the European Parliament is expected to debate the procurement package in the coming weeks. The outcome of these developments will have material effects on model procurement, vendor risk planning, and compliance requirements for organizations operating across borders.
Anthropic, an American AI startup founded in 2021, launched Fable 5 and Mythos 5 on the very day the directive arrived. In a statement released that afternoon, the company explained that the letter “did not provide specific details of the national security concern” and that compliance meant disabling both models for all customers. Anthropic also noted that U.S. officials had become aware of a method for bypassing, or “jailbreaking,” Fable 5. The firm reviewed the technique, disagreed with the finding, and warned that applying such a standard across the industry would “essentially halt all new model deployments for all frontier model providers.”
The ban sent shockwaves through the European Union, where researchers and businesses had already integrated the models into academic and commercial workflows. Earlier in June, Anthropic had granted the European Union’s cybersecurity agency ENISA access to Mythos under the Glasswing project—a partnership now at odds with the sudden suspension.
In response, European Commission Executive Vice‑President for Tech Sovereignty, Security and Democracy Henna Virkkunen posted on social media on June 15 that “Europe represents an economic opportunity, not a security risk. We are and will remain a trusted partner. This is why we need to cooperate on emerging, powerful AI models. This is a shared global challenge.” The comment followed the Commission’s announcement that it was assessing the practical impact of the directive on EU users and cautioned that measures to address risks from advanced AI should not be discriminatory against partners.
Virkkunen said the episode “underlines the need for Europe’s technological sovereignty” and urged swift adoption of a package of measures announced earlier in June. The package includes procurement preferences for European firms in sensitive public‑cloud and AI contracts. It will become law only after negotiation with the European Parliament and member states. The European Union currently sources more than 80 percent of its digital products, services, infrastructure and intellectual property from outside the bloc—a fact that has amplified concerns about digital dependence.
The directive raises several questions for stakeholders. Observers note that the U.S. government has not yet provided technical details backing the order, and it remains unclear whether Anthropic will restore access once the directive is lifted. The European Parliament’s speed in moving on the proposed procurement measures will also influence EU procurement strategy. Additionally, it remains to be seen whether other frontier model providers will face comparable export‑control orders.
The incident underscores the growing regulatory friction between U.S. export controls and the global AI ecosystem. It also highlights the EU’s push for greater technological sovereignty, a policy direction that could reshape how European institutions and enterprises manage vendor risk and compliance across jurisdictions.
At present, Anthropic has not announced a timeline for reinstating Fable 5 or Mythos 5. The European Commission continues to monitor the situation, and the European Parliament is expected to debate the procurement package in the coming weeks. The outcome of these developments will have material effects on model procurement, vendor risk planning, and compliance requirements for organizations operating across borders.
