Anthropic Blocks Access to Fable 5 and Mythos 5 Following Government Order

Anthropic, a leading artificial intelligence safety and research company, announced Friday it has temporarily suspended access to its Fable 5 and Mythos 5 AI models. This decisive action, taken at the behest of the U.S. government, is in response to an export control directive citing national security concerns. The company confirmed it received an order at 5:21 p.m. ET, mandating the immediate cessation of all access to these specific models for any foreign national, regardless of their location or employment status with Anthropic.

To ensure full compliance with the directive, Anthropic opted for a comprehensive suspension, affecting all its customers. The company has assured that its other AI models remain unaffected by this development.

This abrupt move follows closely on the heels of Anthropic’s recent unveiling of Fable 5 and Mythos 5. These models were prominently marketed as groundbreaking, demonstrating state-of-the-art performance across a variety of industry benchmarks. Fable 5, in particular, represented a significant public release of Anthropic’s most advanced offering to date, incorporating novel safeguards designed to mitigate responses in sensitive, high-risk domains. The underlying architecture of these new models builds upon the capabilities demonstrated by the Claude Mythos Preview, which had previously garnered considerable attention from both Wall Street and government agencies for its advanced cybersecurity functionalities.

The company had initially indicated that Claude Mythos Preview would not be widely available, with its rollout restricted to a select group of enterprises as part of Project Glasswing, an initiative focused on enhancing cybersecurity. The latest directive, however, brings a broader AI offering under government scrutiny.

In its official statement, Anthropic acknowledged the government’s directive but noted the absence of specific details regarding the nature of the national security concerns. The company expressed its regrets to its customers for any disruption caused by this unforeseen suspension.

“As we have publicly stated, we believe the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts,” Anthropic stated. “This action does not adhere to those principles.” This sentiment underscores a growing tension between the rapid advancement of AI technologies and the evolving landscape of national security regulations.

This directive marks a notable escalation in Anthropic’s engagements with U.S. governmental bodies. It follows a high-profile dispute with the Department of Defense earlier this year, which saw the agency designate Anthropic as a supply chain risk. This classification, typically reserved for entities perceived as threats to U.S. national security, mandates that defense contractors certify the non-use of Anthropic’s Claude models in their military-related work. Anthropic has since initiated legal proceedings against the administration to challenge this designation, a lawsuit that remains ongoing.

The broader implications of this government intervention are significant. It highlights the complex geopolitical considerations surrounding advanced AI development, particularly concerning the potential for dual-use technologies. For AI developers, navigating the intricate web of export controls and national security directives is becoming an increasingly critical aspect of market strategy and global expansion. The incident raises pertinent questions about the balance between fostering innovation in a highly competitive AI race and safeguarding national interests. As the U.S. government seeks to maintain its technological edge, such measures could reshape the international distribution and access to cutting-edge AI capabilities, potentially influencing research collaborations and market dynamics worldwide. The long-term impact on the pace of AI development and its accessibility to both allies and potential adversaries remains a subject of intense scrutiny.