China Accesses Mythos AI: US Fears Cybersecurity Threat

The Trump administration ordered Anthropic to cut off global access to its Mythos and Fable 5 AI systems on June 13, 2026. This action followed intelligence suggesting a China-affiliated entity may have gained unauthorized access to Mythos. American security officials are concerned China could replicate the model's capabilities using "distillation" techniques. Mythos, released in April 2026, is a cutting-edge cybersecurity AI capable of discovering and weaponizing unknown software weaknesses. It identified thousands of critical zero-day vulnerabilities and could condense months of penetration testing into minutes. Anthropic maintains the White House discussions focused on alleged compromises of Fable 5, the consumer version of Mythos, rather than Chinese penetration. Cryptocurrency exchanges, including Coinbase, lost access to defensive AI capabilities provided through Project Glasswing as a result of these events. This situation highlights the significant national security implications of advanced AI technology.