California AB 412 (Generative artificial intelligence: training data: copyrighted materials.)

Requires developers of generative AI (GenAI) models to document and disclose copyright-protected materials used for training, enable rights owners to request this information, and comply within 30 days. Exempts noncommercial research and models using publicly available data or owned materials.

  • Authority: California
  • Parent authority: State governments
  • Jurisdiction: United States
  • Collections: U.S. state and local documents
  • Most recent activity: Proposed
  • Most recent activity date: 2025-07-09

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

  • Governance failure
  • Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

  • Power centralization and unfair distribution
  • Economic and cultural devaluation of human effort

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has minimal coverage of risk domains, primarily addressing intellectual property and transparency issues related to AI training data. It touches on governance failure (6.5) through its regulatory framework and has implicit connections to power centralization (6.1) through copyright protection mechanisms. The document does not substantially address most AI risk domains as it focuses narrowly on copyright documentation and disclosure requirements.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

  • Information
  • Arts, Entertainment & Recreation

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

  • Scientific Research & Development Services
  • Educational Services
  • Public Administration (excl. National Security)

All other sectors had no coverage in this document.

Sectors Summary

This document primarily governs the Information sector, specifically AI developers and technology companies that create generative AI models. It also has significant implications for the Arts, Entertainment, and Recreation sector through its protection of copyrighted creative works. The law applies cross-sector to any developer using GenAI commercially in California or making it available to Californians.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding California state statute with mandatory obligations, enforcement mechanisms including civil penalties, and specific compliance requirements for GenAI developers.

Actors

Proposer

  • Entity Types: AI_Governance_Actor
  • Summary: The document is a California state bill enacted by the California legislature, as indicated by the opening phrase 'The people of the State of California do enact as follows'.

Target

  • Entity Types: AI_Developer
  • Summary: The law explicitly targets 'developers' of GenAI models, defined as businesses, persons, partnerships, corporations, or other entities that design, code, produce, or substantially modify GenAI models and either use them commercially in California or make them available to Californians.

Enforcer

  • Entity Types: Affected_Stakeholder
  • Summary: The law establishes a private right of action allowing rights owners (copyright holders) to bring civil actions against developers for non-compliance. There is no designated government enforcement agency; enforcement is through the court system via private litigation.

Monitor

  • Entity Types: Affected_Stakeholder
  • Summary: Rights owners are empowered to monitor compliance by submitting requests for information about developers' use of covered materials. The law provides them with mechanisms to request documentation and assess whether their copyrighted materials were used in training.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

  • Collect and Process Data
  • Build and Use Model
  • Deploy
  • Operate and Monitor

AI Lifecycle Summary

The document primarily covers the 'Collect and Process Data' and 'Build and Use Model' stages of the AI lifecycle, focusing on documentation and disclosure of copyrighted training data. It also addresses 'Deploy' through requirements for ongoing availability of information mechanisms, and 'Operate and Monitor' through retention requirements and ongoing obligations while models are in commercial use.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

  • AI Models
  • AI Systems
  • Generative AI
  • Open Weight / Open Source

Technical Scope Summary

The document explicitly defines and covers generative artificial intelligence (GenAI) and AI models. It does not mention frontier AI, general purpose AI, task-specific AI, foundation models, predictive AI, or specific compute thresholds. It implicitly addresses open-weight models through exemptions and does not distinguish between different AI model types beyond generative AI.

https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260AB412
https://agora.eto.tech/instrument/4979
Explore in the Document View

Document data from AGORA, CSET Emerging Technology Observatory.