Source: Almanac News

Almanac News highlights the commencement of a new California law requiring major technology companies to disclose how they manage catastrophic risks stemming from artificial intelligence. Senate Bill 53, authored by Senator Scott Wiener, mandates that developers of frontier models report critical safety incidents to the state within strict timelines, including a 24-hour window for imminent threats of death or injury. The law defines catastrophic risks as events resulting in more than 50 casualties or significant financial damage exceeding $1 billion. This regulatory shift ensures that the public and lawmakers are informed about potential AI-induced hazards, such as the loss of human control over autonomous systems.

In addition to reporting, the law requires developers to publish frameworks on their websites describing their adherence to international safety standards. Companies that fail to comply with these transparency measures may face substantial fines of up to $1 million per violation. The legislation also provides a mechanism for members of the public and company employees to report risks directly to the Office of Emergency Services. By focusing on accountability and information sharing, California aims to lead the nation in mitigating the most extreme risks of AI while continuing to foster its dominant tech industry.