US government allows Anthropic to release Claude Mythos 5 to trusted partners

What happened

On June 26 the US Commerce Department permitted Anthropic to release its most capable cyber model, Claude Mythos 5, to more than 100 trusted partners after two weeks of export restrictions.

Context and impact

Mythos had been blocked following warnings from Amazon and others that it could be jailbroken and misused. Anthropic committed to work with the US government on standards and protocols for releases. The decision came the same day OpenAI released GPT-5.6 to vetted partners — placing both labs in the same federal review regime.

Details

• Commerce Secretary Howard Lutnick wrote to Anthropic chief compute officer Tom Brown
• Mythos can be deployed to a small group of cyber defenders and infrastructure providers
• General Joshua Rudd told a June Senate hearing Mythos 'broke into almost all of our classified systems in hours'
• Access includes some Fortune 500 companies and government agencies
• Decision follows Project Glasswing evaluations