AI labs agree on safety eval standards

xAI, OpenAI, and Anthropic have co-signed a shared standard for third-party AI safety evaluators.

Also today

Today's top five AI stories, one line each, at 7am Eastern. Free. One-click unsubscribe.