Anthropic Limits Release of New Mythos Model Citing Security Concerns

AI lab Anthropic has restricted the release of its newest large language model, Mythos, citing concerns that the system is too capable of identifying security vulnerabilities in widely used software. The company said it limited Mythos availability this week because the model can spot exploits in code relied upon by users globally. The decision has prompted debate about whether Anthropic's reasoning prioritizes public safety or the lab's own interests. Security researchers have long warned that advanced AI systems could accelerate vulnerability discovery and exploitation. Anthropic's move represents one of the first cases of a commercial AI lab withholding a model specifically for security reasons. Mythos is the latest in Anthropic's Claude family of AI models. The company has previously emphasized "AI safety" as a core focus, distinguishing itself from competitors in marketing materials. The incident highlights ongoing tensions in the AI industry between rapid model deployment and potential risks to software infrastructure security.