Multistakeholder regulatory organizations.

Senator McNerney's proposal to establish Multistakeholder Regulatory Organizations (MROs) introduces a new framework for artificial intelligence oversight in California through private-sector certification bodies. Under the legislation, the Attorney General would designate qualified private entities as MROs to evaluate and certify AI models and applications based on safety standards and risk mitigation practices.

The bill outlines specific requirements for MRO designation, including comprehensive plans for auditing AI systems, addressing high-impact risks like cybersecurity and malign persuasion, and maintaining independence from the AI industry. Designated MROs would serve three-year terms and must submit annual reports to the Legislature and Attorney General detailing certification activities, risk assessments, and mitigation measures. The Attorney General retains authority to revoke MRO status for entities that fail to maintain standards or whose certified AI systems cause harm.

For AI developers, MRO certification provides an affirmative defense against civil claims for personal injury or property damage caused by their AI systems, except in cases of intentional misconduct. The certification process requires developers to demonstrate heightened care and adherence to best practices through pre-deployment and post-deployment evaluations. MROs must track and report aggregated data on AI capabilities, observed risks and benefits, and the effectiveness of existing safety measures.

The legislation positions MROs as intermediaries between government oversight and industry innovation, creating a mechanism for establishing and enforcing AI safety standards through private-sector expertise while maintaining public accountability through state supervision and reporting requirements.