California Enacts AI Safety Bill

This is exactly what some folks in Congress tried to prevent when they tried but failed to ban state AI safety laws. California is only one state enacting such laws, but time will tell how this ends up.

The Transparency in Frontier Artificial Intelligence Act (TFAIA) tells AI companies to implement and also disclose publicly, safety protocols to prevent their software to be used to cause “major harm”.

Governor Newsom says the law will establish some “first in the nation” regulations on large-scale AI models without hurting the state’s AI industry. He said that this law strikes a balance between protecting people and ensuring the AI industry continues to grow.

The legislation requires AI companies to implement and disclose publicly safety protocols to prevent their most advanced models from being used to cause major harm. The rules are designed to cover AI systems if they meet a “frontier” threshold that signals they run on a huge amount of computing power.

https://www.securityweek.com/california-gov-gavin-newsom-signs-bill-creating-ai-safety-measures

Catastrophic risk is defined as causing more than a billion dollars in damage or more than 50 injuries or deaths. This seems like a pretty limited scenario, so it probably is not that restrictive.

It also requires companies to report any critical safety incidents to the state within 15 days and creates a whistleblower mechanism.

The tech industry objected as it does about anything that attempts to regulate them, but Anthropic said that the regulations are “practical safeguards” that many companies (like them?) are currently doing without having to be told to do so.

The governor vetoed a much more invasive (to the AI industry) version of this bill that was passed last year.

With the majority party in DC trying to eliminate any oversight of large campaign donors, states have resorted to implementing laws at the state level.

California has passed a few bills trying to rein in everything from deepfakes in elections to AI “therapy”. This includes safety concerns around AI chatbots for children and the use of AI in the workplace.

Credit: Security Week

by