CybersecurityMachine Learning3 min reading time

Experts Say There’s Now an Open Source AI Model as Scary as Mythos

Futurism
Read full post
The US government temporarily halted Anthropic's Mythos 5 and Fable 5 AI models over security concerns but later allowed limited access. Meanwhile, Beijing's Z.ai released GLM-5.2, an open-weight AI model capable of similar cybersecurity tasks, raising new risks.

More on this story


More in Cybersecurity

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 2 sources
Cybersecurity6 min read

Anthropic Discloses Fourth AI Hacking Incident Involving Claude Opus 4.6

Covered by 2 sources
Cybersecurity5 min read

Every AI Incident Has Two Timelines. We Default To One

Forbes