Anthropic Granted Approval To Release Claude Mythos 5 To 100+ U.S. Organizations The Trump administration has authorized artificial intelligence company Anthropic to restore limited access to its flagship Mythos 5 model for a select group of approximately 100 U.S. companies and federal agencies, easing restrictions imposed earlier this month as part of an unprecedented national security review of frontier AI systems. The decision, communicated in a letter from Commerce Secretary Howard Lutnick, represents the first significant breakthrough after nearly two weeks of negotiations between federal officials and Anthropic. While the approval allows certain trusted organizations to begin using Mythos 5 again, the government has stopped short of lifting restrictions on Anthropic's separate Fable 5 model, which remains unavailable pending additional security reviews. The move marks another milestone in the rapidly evolving relationship between Washington and America's leading AI developers, highlighting the growing willingness of the federal government to directly intervene in the release of advanced artificial intelligence systems that officials believe could pose cybersecurity or national security risks. Limited Approval After Intense Negotiations According to multiple reports, Commerce Secretary Howard Lutnick concluded that sufficient safeguards had been implemented to permit Mythos 5 to return under tightly controlled conditions. Rather than restoring public availability, the authorization limits deployment to a carefully selected network of trusted organizations that includes Fortune 500 companies, critical infrastructure operators, cybersecurity partners and federal agencies. Reports indicate many approved organizations participate in Anthropic's Project Glasswing initiative, a program focused on applying advanced AI to defensive cybersecurity operations. The Commerce Department's approval is viewed as a compromise between maintaining national security controls while avoiding a prolonged disruption for organizations that rely on cutting-edge AI tools for software security, vulnerability detection and cyber defense. Government officials emphasized that the authorization could be revised if new security concerns emerge, underscoring