A pair of California state lawmakers has partnered on legislation that would establish safety standards for independent, third-party assessments of artificial intelligence systems and models.
State Sen. Jerry McNerney, D-Pleasanton, and Assemblymember Rebecca Bauer-Kahan, D-Orinda, introduced separate, but complementary bills to regulate AI.
McNerney's
Bauer-Kahan is planning to amend her bill,
"AI has tremendous potential to improve our lives, but without sufficient guardrails, it also poses significant risks," McNerney said. "Under SB 813, California will become the first state to create voluntary standards for the responsible development of AI."
The legislation being
California is the global leader in artificial intelligence innovation and is home to the most significant concentration of AI companies, researchers, and developers in the world, according to the lawmakers.
While artificial intelligence has the potential to drive economic growth, improve public services, advance scientific discovery, and enhance the quality of life for Californians, it's also advancing rapidly and the technology presents significant downsides, including psychological harm to children, inherent bias, cyber fraud, election tampering, misinformation, deepfakes, fake news, and widespread job displacement, according to the lawmakers.
SB 813's AI framework is based on standards models that have been successful in providing benefits for the public in other industries, from electrical and mechanical standards to LEED standards for green buildings, they said.
Under SB 813, the AI Standards and Safety Commission would create standards and can designate working groups composed of experts and stakeholders who can facilitate the development of practical, evidence-based standards that evolve alongside advances in technology. The standards will then be published on the commission's website.
AB 1405 would establish the first-of-its-kind registry and standards for AI auditors. The bill would require independence and rigor in audits and begin to establish an ecosystem for third-party verifiers to oversee the safety of AI systems.
The bill will also require that registered AI auditors meet clear standards for ethical practice and financial independence. This will give Californians confidence that AI audits are conducted with integrity, and the state is taking critical steps to protect them from the risks posed by AI systems.
Good AI policy requires independent verification of safety and AB 1405 would take a significant step in creating the ecosystem for independent verification to become a reality, McNerny said.
By working together, "we are building a framework where California leads the way in creating a system of independent, verified auditors and known standards that protect our communities from AI harms and unchecked power in the hands of a few companies, Bauer-Kahan said.
McNerney's bill has been modified so that the two bills work together. The changes to AB 1405, which has been sitting in the Senate Appropriations Committee's suspense file since August, are planned to be made in the weeks ahead.
SB 813 passed the Senate in January, and was referred to the Assembly Privacy and Consumer Protection Committee, which is chaired by Bauer-Kahan.
