The United States government has announced a partial reversal of its previous aggressive block on Anthropic's flagship artificial intelligence models. This move grants authorization for a select group of American cybersecurity firms and critical infrastructure providers to deploy the advanced Claude Mythos 5 model.
This surprising de-escalation comes just two weeks after the Trump administration abruptly invoked export control authorities. The initial ban forced Anthropic to freeze access to its two most advanced models, Mythos 5 and Fable 5, citing severe national security and potential 'jailbreak' concerns.
Controlled Release for Vetted Organizations
Commerce Secretary Howard Lutnick communicated the decision to Anthropic via a private letter, confirming that intensive negotiations had paved the way for a controlled deployment under a new, rapidly forming regulatory framework. Lutnick stated, "I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model," acknowledging significant progress in bilateral discussions.
The initial re-release will bypass standard export license requirements for approximately 100 heavily vetted organizations. These entities are primarily members of Anthropic’s defensive security initiative, Project Glasswing. This allows them and their foreign national employees immediate access to the system.
Concerns Over AI Exploit Capabilities
The original June 12 shutdown sent shockwaves through the tech sector, occurring just hours after Anthropic officially launched the Mythos-class models. During internal trials, Mythos Preview reportedly demonstrated unprecedented, autonomous multi-step exploit capabilities. This included uncovering a 27-year-old vulnerability in OpenBSD’s TCP stack and engineering functional exploits for Mozilla’s Firefox browser at a rate 90 times faster than prior generations.
Washington's intervention was prompted by fears that such advanced capabilities could be weaponized by foreign adversaries, such as China or Russia, to compromise global banking and critical infrastructure networks.
Balancing Innovation and Security
The administration's swift policy adjustment underscores the high-stakes balancing act between fostering American AI leadership and safeguarding national security. This decision notably coincided with OpenAI’s concurrent release of its own latest model, GPT-5.6, which is also undergoing a restricted, government-validated rollout.
Commerce Department spokesperson Benno Kass remarked, "In just two weeks, we have worked diligently to ensure America remains the global leader in AI while safeguarding our security."
While this limited authorization marks an engineering victory for Anthropic, the broader commercial landscape for its models remains fractured. The government’s directive is notably silent on Fable 5, the consumer-facing version of the model that had briefly been accessible to general subscribers before the ban. For now, non-approved enterprises, global allies, and everyday consumers remain uncertain about when, or if, they will gain access to the new architecture.
Anthropic has confirmed its active cooperation with federal officials to design long-term deployment standards. The company stated, "We're restoring access for these organizations quickly, and we're continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again."