Official name: California AB 316 (Artificial intelligence: Defenses 2025)
Defines "artificial intelligence" as a system influencing environments. Prohibits defendants from claiming AI autonomously caused harm in legal actions. Allows defendants to present other defenses, evidence of causation, foreseeability, and comparative fault of other parties.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding California state statute that creates mandatory legal obligations regarding liability defenses in civil actions involving AI. It uses mandatory language ('shall not be a defense') and establishes enforceable legal requirements through the court system.
The document has minimal coverage of risk domains, with primary focus on legal liability frameworks rather than specific AI risks. It addresses legal accountability mechanisms (6.5 Governance Failure) and implicitly touches on system safety through its prohibition of autonomy-based defenses (7.1, 7.3). Coverage is limited to 2-3 subdomains with minimal detail.
This statute applies broadly across all sectors where AI systems are developed, modified, or used, as it establishes general liability rules for civil actions involving AI-caused harms. No specific sectors are explicitly mentioned or excluded, making it a cross-sectoral legal framework applicable to any industry using AI.
The document does not focus on specific AI lifecycle stages but rather establishes legal liability rules that apply after AI systems have been deployed and allegedly caused harm. It implicitly covers the Deploy and Operate and Monitor stages by addressing post-deployment legal accountability.
The document explicitly defines and covers 'artificial intelligence' broadly as engineered or machine-based systems that can infer from inputs to generate outputs influencing environments. It does not distinguish between different types of AI (frontier, general purpose, task-specific, etc.) or mention compute thresholds or model architectures.
California State Legislature; The People of the State of California
The document is enacted by the California state legislature as indicated by the formal enactment clause and addition to the California Civil Code.
California courts; California judicial system
Enforcement occurs through the California court system in civil actions, where judges will apply this statute to prohibit defendants from asserting AI autonomy as a defense.
No explicit monitoring body is designated. Monitoring would occur implicitly through the California judicial system as cases are adjudicated and legal precedents are established.
The statute explicitly targets defendants who developed, modified, or used artificial intelligence in civil actions where AI allegedly caused harm. This encompasses developers, deployers, and users of AI systems.
3 subdomains (3 Minimal)