Anthropic Receives Approval To Release Claude Mythos AI Model To Trusted Partners - NDTV

The US government has given Anthropic the green light to release its powerful Claude Mythos 5 AI model to over 100 'trusted partners', including major companies and government agencies, after a tense two-week standoff over national security concerns. This move partially reverses an earlier government order that had suspended global access to Mythos 5 due to fears its advanced cybersecurity capabilities could be misused by foreign adversaries. The drama unfolded after the Trump administration, initially leaning towards minimal AI regulation, abruptly ordered Anthropic to halt access to Mythos 5 and its related model, Fable 5, amidst warnings from companies like Amazon about the models being 'jailbroken' to exploit software vulnerabilities. Mythos 5 is a frontier AI, capable of advanced cybersecurity tasks, autonomous coding, and identifying security flaws, even conducting multi-stage attacks in controlled environments, which raised alarms in Washington. This intervention aligns with a recent Executive Order signed on June 2, 2026, which established a voluntary framework for AI developers to share powerful models with the federal government before wider release, signaling a significant shift in US AI policy towards national security. While Anthropic can now deploy Mythos 5 for critical infrastructure defense, access to the more broadly intended Fable 5 model remains restricted, though discussions are ongoing for its future release. This government oversight isn't just limited to Anthropic; rival OpenAI also announced a limited release of its new GPT-5.6 models to government-approved partners on the same day, indicating a tightening regulatory environment for frontier AI. The evolving relationship between powerful AI developers and national security concerns is clearly a space to watch closely.