California AB 2839 (2024)

• Authority: California
• Parent authority: State governments
• Jurisdiction: United States
• Collections: U.S. state and local documents
• Most recent activity: Enacted
• Most recent activity date: 2024-09-17

Risk subdomain coverage

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

• False or misleading information
• Pollution of information ecosystem
• Disinformation, surveillance, and influence at scale
• Fraud, scams, and targeted manipulation

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

• AI system security vulnerabilities
• Overreliance and unsafe use
• Loss of human agency and autonomy
• AI possessing dangerous capabilities

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 6-8 subdomains, with strong focus on malicious actors (4.1 Disinformation, surveillance), misinformation (3.1 False information, 3.2 Information pollution), AI system security (2.2), and human-computer interaction (5.1 Overreliance). Coverage is concentrated in election integrity, misinformation prevention, and malicious use domains.

Sector coverage

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

• Information
• Public Administration (excl. National Security)

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

• Professional & Technical Services
• Arts, Entertainment & Recreation

All other sectors had no coverage in this document.

Sectors Summary

The document primarily governs the Information sector (broadcasting, telecommunications, internet platforms) and Public Administration (elections officials, electoral processes). It also has significant implications for Professional and Technical Services (political consultants, campaign services) and Arts, Entertainment, and Recreation (media production).

Legislative Status

Hard Law

Legislative Status Summary

This is a binding California statute with mandatory prohibitions, civil enforcement mechanisms including injunctive relief and damages, and explicit legal penalties for violations.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Summary: The document is a California state statute enacted by the California Legislature, as indicated by the opening phrase and the urgency statute designation.

Target

• Entity Types: AI_Developer, AI_Deployer, Affected_Stakeholder
• Summary: The statute explicitly targets any person, committee, or other entity that distributes materially deceptive AI-generated content related to elections, with specific provisions addressing broadcasters, internet platforms, and publishers.

Enforcer

• Entity Types: AI_Governance_Actor, Affected_Stakeholder
• Summary: The statute provides for enforcement through civil actions in California courts, which can be brought by recipients, candidates, committees, or elections officials seeking injunctive relief or damages.

Monitor

• Summary: The statute does not establish a specific monitoring body or oversight mechanism. Enforcement relies on private civil actions rather than proactive government monitoring.

AI Lifecycle Stages

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

• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document focuses primarily on the Deploy and Operate and Monitor stages, as it regulates the distribution and use of AI-generated content in election contexts. It does not address earlier lifecycle stages such as planning, data collection, or model development.

Technical Scope

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

• Generative AI

Technical Scope Summary

The document explicitly addresses AI-generated content, deepfakes, and generative AI. It defines deepfakes and materially deceptive content but does not distinguish between different types of AI models (foundation models, general purpose AI, etc.) or specify compute thresholds. The focus is on the output (deceptive content) rather than the underlying AI technology.