Anthropic restricts access to Claude Mythos AI for identifying security vulnerabilities
Anthropic
Anthropic has limited access to Claude Mythos, an AI that identifies security flaws. The model was found to be effective too quickly for safe deployment, raising concerns about dual-use misuse. Access is now restricted.
Anthropic has limited access to its Claude Mythos AI model, which is designed to identify security vulnerabilities. The restriction was implemented because the model proved to be effective too quickly, potentially posing risks of dual-use misuse where it could be used for malicious purposes. As a result, access to Claude Mythos is now restricted to prevent unintended negative consequences.
Source: Anthropic (GNews) —
original