AI safety hinges on control, not raw smarts

A new framework argues AI governance should restrict what agents can access and control, not cap their intelligence.

Also today

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