The US government has reversed its decision to block access to two of Anthropic's most powerful artificial intelligence models, Fable 5 and Mythos 5, after the company addressed security vulnerabilities. The San Francisco-based AI lab announced on Tuesday that it had received notice from the Department of Commerce lifting export controls, allowing it to restore global access to the systems.
Anthropic, best known for its Claude chatbot, had been forced to cut off access to these models on 12 June after US authorities discovered weaknesses in safeguards designed to prevent misuse. The abrupt restriction highlighted Washington's growing scrutiny of advanced AI technologies, a trend that has also affected competitors like OpenAI, which limited the release of its GPT-5.6 model to approved partners.
In a post on X, Anthropic stated: "We've received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5," adding, "We'll begin restoring access tomorrow." The company had previously received authorization for a small group of American cybersecurity firms to access Mythos 5 just four days before the broader lifting.
Security Concerns Addressed
The restrictions were lifted after a letter, seen by US media, indicated that the Trump administration was satisfied that Anthropic had "taken steps in close coordination with the U.S. government to address the risks associated with Claude Mythos 5 and Claude Fable 5." The process underscores the delicate balance between fostering innovation and ensuring national security in the rapidly evolving AI landscape.
For European observers, this episode raises questions about the continent's reliance on US-developed AI models. While the EU has been proactive in regulating AI through the AI Act, the bloc's tech ecosystem remains heavily dependent on American cloud infrastructure and foundational models. The temporary block on Anthropic's models could have disrupted European businesses and researchers who rely on these tools for tasks ranging from data analysis to content generation.
European policymakers have been pushing for greater digital sovereignty, with initiatives like the European High-Performance Computing Joint Undertaking and investments in homegrown AI startups. However, the gap between US and European AI capabilities remains significant. The US government's ability to unilaterally restrict access to cutting-edge models serves as a reminder of the strategic importance of developing independent AI capacity within Europe.
Anthropic's CEO Sam Altman acknowledged that the process was not optimal, but the company's compliance with US government requests reflects the broader trend of AI labs cooperating with authorities to mitigate risks. The incident also highlights the challenges of governing AI in a globalized world, where models developed in one country can have far-reaching implications.
As Europe continues to shape its AI strategy, the continent will need to balance innovation with regulation, ensuring that its citizens and businesses can access the benefits of advanced AI without being overly dependent on external actors. The lifting of export controls on Anthropic's models is a positive development for global AI users, but it also underscores the need for Europe to strengthen its own technological foundations.
