According to a report from Fortune, the White House ordered the shutdown of Anthropic's Mythos AI model after Amazon issued a warning that the model posed significant national security risks. The decision, made in early June 2026, followed Amazon's assessment that Mythos could be used for malicious purposes, including cyberattacks or disinformation campaigns.
Anthropic, an AI safety company, had been developing Mythos as a large language model with advanced capabilities. However, Amazon, which is a major investor in Anthropic, raised concerns with federal officials after internal testing revealed potential vulnerabilities. The White House's National Security Council then intervened, citing the need to prevent potential threats to critical infrastructure.
Neither Anthropic nor Amazon have publicly commented on the specifics of the warning. The shutdown has sparked debate among AI experts about the balance between innovation and security. Some argue that the move sets a precedent for government oversight of advanced AI models, while others worry it could stifle progress in the field.
As of June 15, 2026, the Mythos model remains offline, and Anthropic is reportedly working with regulators to address the security concerns. The incident highlights the growing tension between tech companies and the U.S. government over AI regulation.
