AI Model Safety Standards: Five Labs Sign On Ahead of August 1 DeadlineNew AI model safety standards from the White House add a shared jailbreak scoring scale and a 30-day pre-release review for frontier models.7 min00