The California Assembly Privacy and Consumer Protection Committee convened on May 27, 2025, to discuss critical issues surrounding the regulation of artificial intelligence (AI) and the role of third-party evaluations in ensuring safety and security. The meeting highlighted the challenges of establishing effective oversight in a rapidly evolving technological landscape.

A significant point of discussion was the need for a threshold system to differentiate between smaller companies and larger entities regarding regulatory requirements. Participants emphasized that smaller companies should be protected from excessive regulatory burdens, suggesting that those below a certain threshold should not be required to undergo extensive testing. In contrast, larger companies would need to comply with more rigorous evaluations, which could adapt as technology advances.

The conversation also addressed the current inadequacies in the ecosystem for third-party evaluations. Experts noted that while third-party auditing is essential for accountability, the market for such services is still developing. There was a consensus that transparency is crucial, and establishing a regulatory framework could stimulate the growth of this market. The idea is that if the government signals the need for independent evaluations, companies will emerge to fill that gap.

Additionally, the committee discussed the importance of flexibility in the evaluation process. It was suggested that if companies cannot find suitable third-party evaluators, they should be allowed to conduct their assessments, provided they justify their decisions to the government. This approach aims to ensure that evaluations are conducted without stifling innovation.

The meeting concluded with a recognition of the sensitive nature of AI testing, particularly concerning potential misuse. Participants advocated for the establishment of secure facilities for testing AI models, especially those that could pose risks if mishandled. The need for policy incentives to encourage the development of such secure environments was highlighted as a critical step forward.

Overall, the discussions underscored the importance of balancing regulatory oversight with the need for innovation in the AI sector, as California continues to navigate its role as a leader in technology and consumer protection.