Anthropic: American-Controlled AI Worries Europe

The Thunderclap of American Restrictions

The recent export controls imposed by the U.S. government against Anthropic have turned a latent political concern into a concrete reality. Since June 13, 2026, a directive has taken offline the company's two most advanced AI models, Fable 5 and Mythos 5, for all foreign users, including temporarily for non-American employees of Anthropic. This decision has triggered a shockwave across Europe and Canada, raising questions about who truly holds power over crucial AI technologies for the world.

The speed at which these events unfolded has been striking, and the international reaction was swift. The U.S. government, citing national security reasons, issued an export control directive that suspended all access to Fable 5 and Mythos 5 for any foreign national, whether inside or outside the United States.

A Launch Followed by a Quick Shutdown

On June 9, 2026, Anthropic launched its AI models, Claude Fable 5 and Claude Mythos 5, to the public. These models represented the showcase of a new generation of technologies developed under the Project Glasswing program since April. Fable 5, described as a secure Mythos-class model for general use, stood out for its exceptional performance in various fields, such as software engineering and scientific research. Mythos 5, more powerful, remained reserved for selected partners and biology researchers.

Four days after their availability, these models were pulled from the market. Anthropic received an export control directive on June 12 at 5:21 PM Eastern Time, without detailed explanations regarding security concerns. Unable to filter users by nationality in real-time, the company had to "abruptly disable" access for all users to comply with the order issued by Secretary of Commerce Howard Lutnick.

The "Jailbreak" Controversy

Washington's decision was based on national security concerns, particularly a method allowing users to "jailbreak" Fable 5, meaning to bypass its security measures. Anthropic downplayed the severity of this technique, stating that it only allowed limited capability to examine the code and identify errors, a feature that other competing models, such as OpenAI's GPT-5.5, also possess.

However, the U.S. government maintained a stricter stance. David Sacks, co-chair of the President's Council of Advisors on Science and Technology, stated that the administration had asked Dario Amodei, CEO of Anthropic, to fix the vulnerability or withdraw the model from the market. Amodei reportedly refused, leading to the suspension. Sacks criticized Anthropic for downplaying the situation in their public communications.

A Conflict Rooted in the Past

This dispute is not new. Earlier in the year, Anthropic insisted that its technology should not be used for mass surveillance or in autonomous weapon systems, which angered Pentagon chief Pete Hegseth. In response, President Trump ordered all federal agencies to stop using Anthropic's technology, labeling the company a "risk to the national security supply chain." Anthropic challenged this decision in court, warning that it could jeopardize hundreds of millions of dollars in revenue.

This paradoxical situation has led to Anthropic being perceived as too dangerous for use by the U.S. government, while also deemed too risky for international use. Dean Ball, an AI policy expert, described this decision as "caricatural," highlighting the inconsistency of an administration willing to export AI chips to China while banning access to Anthropic's models for allies like Great Britain.

International Reactions and Implications

Internationally, the reaction has gone beyond mere technical debate over the jailbreak. It has highlighted an uncomfortable reality: a tool integrated into businesses and institutions worldwide was disabled by a foreign government in a single afternoon.

The European Commission announced that it is examining the implications of this decision. Thomas Regnier, a spokesperson for the Commission, emphasized that while advanced AI models offer significant benefits, particularly for cybersecurity, they also pose cybersecurity challenges that must be addressed. He insisted that the measures taken should not discriminate against international partners.

European politicians expressed more direct criticisms. In France, the American decision was seen as a catalyst for the geopolitical battle over AI, with calls for Europe not to remain a mere market dependent on foreign technologies. Aura Salla, a Finnish MP, stated that Europe could not continue to rely on access that could be cut off overnight by a foreign government.

Future Perspectives

Anthropic maintains its position, asserting that applying this standard across the industry could hinder the deployment of new models for all advanced technology providers. The path to resolution lies with the Bureau of Industry and Security at the Department of Commerce, where a license is now required for the export, re-export, or domestic transfer of the affected models.

David Sacks summarized the situation by stating that the solution lies in fixing the jailbreak to lift the control. "The ball is in Anthropic's court," he asserted. For foreign governments, the question of the fix is almost secondary. The main lesson drawn from this affair is that access to cutting-edge AI is no longer just a matter of cost or product, but of control. Last week, Washington demonstrated that it holds this control, a reality that has raised concerns in many global capitals.