OpenAI Proposal for Harmonized AI Regulation in California
OpenAI has called on Governor Gavin Newsom to propose a harmonized approach to AI regulation, urging California to align its state-based rules with national and global safety standards. The goal is to prevent a fragmented regulatory environment where inconsistent state laws impede technological progress and national security.
National and Global Alignment of AI Safety Standards
OpenAI recommends that California avoid duplicating or creating inconsistencies between state requirements and the existing safety frameworks established by the US government and its democratic allies. The company argues that a national model for AI safety should take precedence over state-specific rules to ensure consistency.
To achieve this alignment, OpenAI proposes that California treat frontier model developers as compliant with state requirements if they have met one of the following criteria:
- Entered into a safety-oriented agreement with a relevant US federal agency, such as the Center for AI Standards and Innovation (CAISI).
- Signed onto a parallel framework, such as the EU’s AI Code of Practice.
OpenAI has already committed to working with CAISI to evaluate the national security-related capabilities of frontier models.
Protecting Small Developers and Startups
To maintain a vibrant AI ecosystem, OpenAI urges the state to avoid imposing compliance burdens on smaller developers and startups. While large firms can absorb the costs of regulatory compliance, early-stage teams often cannot. OpenAI warns that duplicative state rules could create a regulatory environment similar to the California Environmental Quality Act (CEQA), which the company claims made housing construction significantly harder in the state due to a lack of understanding of broader impacts.
Strategic Imperatives and Global Competitiveness
OpenAI argues that that aligning California with US government standards is a strategic necessity to maintain the US lead over autocratic AI development, specifically citing the AI companies in the People's Republic of China (PRC). The company asserts that that AI companies in the PRC are unlikely to abide by US state laws and would benefit from a patchwork of inconsistent regulations that slow down US competitors.
Economic Impact and Mission
OpenAI notes that the AI sector is already contributing billions to California's budget and driving innovation. The company states that its mission as a nonprofit is to build democratic AI anchored in democratic values and safety standards, ensuring broad access to the technology's benefits for as many people as possible.