Anthropic Mythos 5 Returns After US Government Approval
The United States government has allowed Anthropic to restore limited access to its advanced AI model, Mythos 5. The model will only be available to approved organisations involved in critical infrastructure and cyber defence. The decision reverses a suspension that began on June 12 after officials raised concerns about the model’s cybersecurity capabilities and possible misuse by foreign actors.
The government suspended Mythos 5 and the consumer-facing Fable 5 earlier this month over national security concerns. Officials feared the model could help malicious actors identify weaknesses in banking systems, power grids, and other critical infrastructure. The US Commerce Department also wanted advanced AI developers to share powerful models with federal officials before broad releases.
Anthropic Mythos 5 is now being restored to more than 100 trusted US organisations, including government agencies and Fortune 500 companies. The approved list reportedly includes institutions participating in Anthropic’s Project Glasswing initiative, which focuses on cybersecurity and infrastructure protection. While the decision came after negotiations between Anthropic and the Trump administration. Commerce Secretary Howard Lutnick said the company had implemented additional safeguards and made significant progress in addressing security concerns.
Mythos 5 is considered Anthropic’s most powerful cybersecurity model. The AI can analyse software, identify vulnerabilities, and assist with complex research and coding tasks. Unlike Fable 5, Mythos 5 does not include some of the safety classifiers that automatically refuse risky requests. The model is also designed for advanced reasoning and long-running tasks, making it one of the most capable AI systems currently available.
