Bill > SB420

This bill introduces comprehensive regulations for high-risk automated decision systems (AI systems that significantly impact important life decisions) in California. The legislation requires developers and deployers of such systems to conduct thorough impact assessments before making these systems publicly available or deploying them, with these assessments needing to detail the system's purpose, potential risks of algorithmic discrimination, and safeguards to prevent unfair outcomes. State agencies would be required to obtain and keep these assessments confidential, and developers must provide impact assessments to the Attorney General or Civil Rights Department upon request. The bill establishes that these AI systems should not discriminate based on protected characteristics like race, gender, or disability, and must have mechanisms for individuals to understand how decisions are made and appeal automated decisions. Developers and deployers would face potential civil penalties for violations, ranging from $2,500 to $25,000 depending on company size and the nature of the violation, with an opportunity to cure most violations within 45 days. The legislation applies to AI systems used in critical areas such as employment, education, healthcare, housing, and government services, with exemptions for very small companies (50 or fewer employees) and systems already approved by federal agencies with comparable regulations. The bill ultimately aims to establish a framework that ensures AI technologies are developed and deployed responsibly, with strong protections for individual rights and prevention of potential discriminatory practices.

Budget and Finance, Government Affairs, Justice

https://leginfo.legislature.ca.gov/faces/billStatusClient.xhtml?bill_id=202520260SB420