California SB 942 vs EU AI Act: A Compliance Comparison

Two Frameworks, One Problem

Both California’s SB 942 (AI Transparency Act) and the EU AI Act address the same core problem: AI-generated content circulating without any indication of its origin. But the two frameworks take meaningfully different approaches, and organizations operating in both jurisdictions need to understand where they overlap and where they diverge.

Scope

SB 942 applies to large AI systems that generate content for California users. The threshold is systems with more than one million monthly users. Covered systems must provide disclosure tools that let users identify AI-generated content.

EU AI Act Article 50 applies more broadly to any AI system that generates synthetic content interacting with natural persons, with limited exceptions for clearly labeled satire or artistic contexts. It applies to providers and deployers operating in the EU, regardless of where they’re based.

Technical Requirements

SB 942 requires disclosure mechanisms but does not mandate a specific technical standard. The EU AI Act’s draft Code of Practice explicitly references C2PA as the expected implementation standard, with a dual-layer requirement (metadata + imperceptible watermarking).

Enforcement Timeline

• SB 942: Effective January 1, 2026. Enforcement by California Attorney General.
• EU AI Act Article 50: August 2, 2026 enforcement date.

The Good News: C2PA Covers Both

A C2PA-based provenance infrastructure satisfies both frameworks simultaneously. The metadata and watermarking required for EU compliance exceeds what SB 942 technically demands, meaning EU-compliant implementation is automatically SB 942-compliant as well.

This is the case Limbo makes to global organizations: one infrastructure layer, multi-jurisdictional compliance.