anthropic is currently investigating reports of unauthorized access to its highly advanced AI model, Claude Mythos Preview. This development has rAIsed significant concerns within the cybersecurity community, given the model's potent capabilities.
According to a Bloomberg report, a small group of users on a private online forum successfully accessed Mythos on the Same day Anthropic announced its limited release for testing. The model, described as Anthropic's most powerful to date, can autonomously identify thousands of software vulnerabilities and chain them together to form complex attack vectors capable of compromising major platforms.
Due to these powerful offensive capabilities, Anthropic has not released Mythos to the public. Instead, it was made available exclusively to a select coalition of major industry players under a security initiative called Project Glasswing. The project aims to proACTively secure critical software infrastructure.
Project Glasswing partners include:
Comments & Questions (0)
No comments yet
Be the first to comment!