Anthropic's Mythos AI model accessed by unauthorised users, Bloomberg News reports
The group has been using Mythos regularly since then, though not for cybersecurity purposes, according to the report.
Anthropic investigates Mythos AI breach
- Anthropic investigates report of rogue access to hack-enabling Mythos AI The Guardian —
- Anthropic’s most dangerous AI model just fell into the wrong hands The Verge —
- Anthropic’s Mythos Accessed by Unauthorized Users Bloomberg —
- Anthropic’s Mythos Accessed by Unauthorized Users Bloomberg —
- Anthropic investigating unauthorised access of powerful Mythos AI model Financial Times —