The United States government established reporting mechanisms for artificial intelligence labs to report breaches, hacks, and jailbreaks, though pre-launch model testing is not required.