AI companies face new regulations under proposed California bill

A new bill proposed by California State Senator Scott Wiener could transform Artificial Intelligence development in California, imposing new restrictions on industry leaders like OpenAI and Microsoft.

The bill seeks to implement "common sense safety standards" on companies that develop large AI models, according to a report by Bloomberg. The bill would reportedly mandate that companies reduce potential harm caused by generative AI and guarantee the deactivation of the models if deemed necessary.

Under the bill, companies would also have to report on compliance measures to a new division of  the California Department of Technology called the "Frontier Model Division", as reported by Bloomberg. Failure to comply would then lead to civil penalties and possible lawsuits. There remains no federal legislation regulating AI, and as such states are increasingly enacting their own laws to mitigate the potential adverse impacts of AI.

Tech founders, investors, and employees have reportedly voiced concerns about the potential impact of the bill on their operations. Critics of the measure argue it could impose an unmanageable burden on developers, particularly those in the open-source sector, who would be held accountable for preventing the abuse of their services by malicious users.

Senator Wiener has reportedly responded to these concerns and revised the bill, altering the requirements for covered models, clarifying that open-source developers will not be held accountable for all negative uses of their platforms, and specifying that the deactivation requirements won't affect open-source technology.

The senator has expressed support for AI development, according to Bloomberg, stipulating that AI development should be pursued in a safe and sustainable manner. A state assembly is set to vote on the bill at the end of August, according to Bloomberg, and Senator Wiener remains hopeful about the bill's progress.