California: New AI Rules Challenge Washington

California Imposes Strict Rules for AI

California Governor Gavin Newsom recently signed an executive order that imposes new requirements on companies working with the state. These companies must now implement measures to prevent the misuse of artificial intelligence (AI). AI systems must be designed to avoid generating illegal content, perpetuating harmful biases, and violating civil rights.

Measures Against Misinformation

To combat misinformation, the order requires state agencies to watermark all AI-generated images and videos. This measure aims to ensure that content created by AI is easily identifiable, thereby reducing the risk of information manipulation.

Independent Risk Assessment

A notable aspect of the order is California's response to federal guidelines regarding supply chain risks. For example, the Pentagon has designated Anthropic as a risk, prohibiting government contractors from using its technology for U.S. military work. However, California will conduct its own assessment, which could allow the state to continue collaborating with suppliers that the federal government has excluded.

Development of New Certifications

Within 120 days of the order's signing, California's procurement and technology agencies must propose recommendations for new AI certifications. These certifications will enable companies to demonstrate their compliance with responsible AI practices and public safety.

A Challenge to Federal Policy

This order reinforces California's commitment to charting its own course in AI regulation, independent of the Trump administration, which repeatedly attempted to block state-level AI laws.