Anthropic Continues Limited Access to Mythos AI Model Despite Security Suspension

Despite Washington's order to suspend Mythos 5 and Fable 5 due to national security concerns, a select group of partners are still able to access Mythos Preview, an older version of Anthropic's AI model. Bloomberg reported on June 19, 2026, that these privileged users continue to utilize the technology. The suspension was implemented to address potential security risks associated with the advanced AI. This ongoing access for a few raises questions about the effectiveness and scope of the security measures. The situation highlights the complex balance between AI development and national security imperatives. It remains unclear how long this limited access will persist.

This situation is significant because it involves ongoing access to a potentially sensitive AI technology despite government-imposed security suspensions.