Big scoop from me: Anthropic's Mythos AI model — the cybersecurity model it says is so powerful it can enable dangerous cyberattacks — is being accessed by unauthorized users. Gift link!

→ View original post on X — @rachelmetz