The U.S. government authorizes Anthropic to release its “Mythos” artificial intelligence model to selected U.S. organizations described as “trusted” partners. Multiple reports state that the approval allows Anthropic to provide access to the model under government conditions rather than making it broadly available. Hacker News discussion posts and two news outlets describe the authorization as applying to “trusted” or “trusted partners” within the United States.
The reports do not describe a single, unified list of the approved recipients in the provided excerpts. They also do not detail technical specifications, model capabilities, or whether the government’s approval includes limits on deployment, monitoring, or other compliance requirements beyond the “trusted” designation. Overall, the coverage is consistent that the U.S. has granted permission for a more limited, controlled release of Mythos by Anthropic, contingent on the recipients meeting the government’s “trusted” criteria.
