Bill > AB1018

This bill establishes comprehensive regulations for automated decision systems (ADS) that make or facilitate consequential decisions affecting individuals in areas like employment, education, housing, healthcare, and more. The bill defines an ADS as a computational process using machine learning, artificial intelligence, or data analytics that generates simplified outputs like scores or recommendations to assist or replace human decision-making. Developers of covered ADS must conduct annual impact assessments that evaluate the system's accuracy, potential disparate impacts, and intended uses, and beginning in 2030, must contract with independent third-party auditors. Starting January 1, 2027, deployers of these systems must provide subjects of consequential decisions with detailed disclosures about how the ADS works, what information it uses, and offer opportunities to correct personal information or appeal decisions. The bill allows public entities like the Attorney General to bring civil actions for non-compliance, with potential penalties up to $25,000 per violation. Importantly, the bill aims to increase transparency and accountability in AI-driven decision-making while protecting trade secrets and ensuring that the use of automated systems does not perpetuate discrimination, though compliance with the law does not automatically prove or disprove discriminatory intent.

Budget and Finance, Business and Industry, Justice

https://leginfo.legislature.ca.gov/faces/billStatusClient.xhtml?bill_id=202520260AB1018