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

Beijing-based Z.ai released GLM-5.2, an open-weight AI model matching Anthropic's Mythos 5 in cybersecurity vulnerability detection, now freely downloadable and runnable without vendor oversight—raising stakes for malicious actors. Meanwhile, the US government restricted then selectively re-released Anthropic's frontier models citing jailbreak vulnerabilities.

This Wire brief sits within Fusion42's coverage of AI Frontier Models.