The United States government plans to introduce a new executive order to control artificial intelligence and improve cybersecurity. This order will set up a formal system for the government to check powerful AI technologies before they are made public. The measure comes as there are concerns about advanced AI models and their possible security risks.

The order will ask leading AI companies to share information and give government agencies early access to their latest models for 90 days before release. This review period will help security agencies look for dangers, such as ways to attack or hack important networks.

The new rule will also require higher cybersecurity standards for services like power grids, hospitals, and banks. These kinds of essential services are called critical infrastructure. Government officials say that better protection will help keep the country safe and strong.