Google, Microsoft, and xAI have agreed to share unreleased AI models with the U.S. Department of Commerce to conduct safety checks ensuring these technologies do not pose any threats, as announced by the Center for AI Standards and Innovation (CAISI).