Anthropic Restores Claude Mythos 5 Access Following Government Security Review

The United States government has partially lifted export controls on Anthropic, allowing the AI firm to restore access to its powerful Claude Mythos 5 model for a select list of over 100 trusted American corporate and government partners. Commerce Secretary Howard Lutnick confirmed the decision in a letter to the company, noting that Anthropic has satisfied new government security safeguards and agreed to collaborative protocols for future AI deployments. This move follows a tense two-week suspension period that disrupted operations across the AI sector.

The restrictions were initially imposed by the Trump administration amid concerns regarding the security of AI infrastructure and the potential for model misuse. While Anthropic addressed issues regarding a theoretical jailbreak vulnerability in its publicly available Claude Fable 5 model, the government maintained a blanket block during its initial inquiry. The restoration signals a shift toward a more mediated approval process where government agencies act as gatekeepers for sensitive high-end generative AI technology.

While Claude Mythos 5 is now accessible to vetted institutions, Claude Fable 5 remains offline, with discussions between Anthropic and federal regulators currently underway to determine a recovery timeline. The incident underscores a tightening regulatory environment in the United States, which directly impacts global AI firms relying on US-based data infrastructure. For India, which increasingly integrates these advanced models into its digital ecosystem, the ongoing oversight illustrates the growing geopolitical influence of domestic security mandates on international AI access.