Unauthorized group has gained access to Anthropic's exclusive cyber tool Mythos, report claims
TechCrunch • • business
Key Points:
- A group of unauthorized users reportedly gained access to Anthropic's cybersecurity AI tool, Mythos, through a third-party vendor, according to Bloomberg.
- The group, active on a Discord channel focused on unreleased AI models, accessed Mythos the same day it was publicly announced and has been using it regularly, providing evidence to Bloomberg.
- Anthropic is investigating the breach but has found no evidence that its own systems were impacted; the unauthorized access was facilitated by an employee at a third-party contractor.
- Mythos was released selectively to vendors like Apple under Project Glasswing to prevent misuse, as the tool could potentially be weaponized against enterprise security.
- The unauthorized users are reportedly interested in experimenting with new models rather than causing harm, but the breach raises concerns about the security of Anthropic’s exclusive AI releases.
