Articles
SaferAI's independent evaluation finds Zhipu AI's GLM-5.2 performs near saturation on offensive cybersecurity benchmarks while refusing none of the dangerous tasks it was given. Frontier developers like Anthropic and OpenAI consistently refused the same requests.
Mistral AI released Shieldstral on Monday, a 3-billion-parameter open-weights safety classifier that matches or outperforms guard models up to seven times its size. The model is available under Apache 2.0 and runs on a single 16 GB GPU. What sets Shieldstral apart from typical guardrail models is its approach to content moderation. Rather than baking a fixed taxonomy of harm categories into the model weights — which forces developers to retrain whenever their safety requirements change — Shield
Anthropic's most capable public model is generally available again after a three-week export-control pause. The interesting part isn't the model — it's the classifier stack now wrapped around it.
Years of patient mech-interp work just turned a corner. The result is the most concrete reason for optimism in alignment research in years.
