California’s leading artificial intelligence bills are facing increasing scrutiny as they move closer to becoming law. Two major proposals, one by Sen. Scott Wiener requiring safety testing for large AI models and another by Assemblymember Rebecca Bauer-Kahan aimed at preventing bias in algorithmic decisions, have passed initial votes but now confront major challenges in appropriations committees.
If successful there, they must still secure floor votes and gain the approval of Gov. Gavin Newsom, who is known for his tech-friendly and fiscally cautious approach.
A key issue is the cost of Bauer-Kahan’s Assembly Bill 2930, which the Senate Appropriations Committee estimated could run into billions of dollars, a daunting figure in a year marked by budget deficits. Bauer-Kahan disputes this estimate and is working on cost-cutting amendments to make the bill more viable.
At the same time, Wiener’s bill is meeting resistance from fellow Democrats in California’s House delegation, including Reps. Zoe Lofgren and Ro Khanna, who argue the legislation could harm California’s competitiveness in the tech industry. The concerns echo those of AI companies and investors worried about the potential impact on the industry.
The stakes are high, as the legislation could set new standards for AI regulation in the U.S. Wiener and Bauer-Kahan argue that state-level action is necessary due to the federal government’s inaction on emerging technologies, despite the political risks and economic concerns associated with their bills.