Unauthorized group accesses Anthropic's AI security tool

A report claims an unauthorized group has gained access to Anthropic's exclusive cybersecurity AI tool, Mythos. The tool, designed to bolster enterprise security, could pose a significant risk if misused. The group reportedly accessed Mythos through a third-party vendor, using knowledge of Anthropic's model formatting to locate it online. While the group claims to be interested in exploring new AI models rather than causing harm, Anthropic is investigating the alleged breach. The company stated it has found no evidence of its own systems being impacted by the unauthorized activity.