AI policy explorer

California AI Transparency Act (SB 942)

California AI Transparency Act (SB 942, as amended by AB 853)

Covered providers of generative AI systems must offer a free AI-detection tool and embed disclosures and provenance data in AI-generated content. AB 853 (Chapter 674, Statutes of 2025) moved the operative date to 2 August 2026 and added duties for large online platforms and generative-AI hosting platforms.

California SB 53

California SB 53: Transparency in Frontier Artificial Intelligence Act

SB 53 requires "large frontier developers" (developers of the most compute-intensive models above statutory thresholds) to publish a frontier AI framework describing how they assess and mitigate catastrophic risks, publish transparency reports when deploying new frontier models, report critical safety incidents to the California Office of Emergency Services, and protect employees who report safety concerns. It also directs creation of a public computing cluster ("CalCompute").

California training-data transparency law (AB 2013)

California AB 2013: Generative artificial intelligence: training data transparency (Chapter 817, Statutes of 2024)

Developers of generative AI systems made available to Californians must post documentation of the data used to train them on their websites, including summaries of the datasets, whether they include copyrighted material, whether they were purchased or licensed, and whether they include personal information.

Connecticut online safety and frontier AI law (SB 5)

Connecticut SB 5: An Act Concerning Online Safety (Public Act 26-15)

A broad online-safety act that covers frontier developers, AI companions, AI in employment and synthetic content. Its frontier provisions bar retaliation against whistleblowers and require large frontier developers to set up an anonymous internal reporting process and report to their boards quarterly; no duty to publish a safety framework or transparency report was found.

EU AI Act

Regulation (EU) 2024/1689 — Artificial Intelligence Act

The EU AI Act is a binding regulation that sets rules for developing, placing on the market and using AI systems in the European Union. It bans a small set of practices considered unacceptable, imposes detailed requirements on "high-risk" AI systems used in areas such as employment, education, credit, essential services, law enforcement and safety-critical products, requires transparency for chatbots and synthetic content, and creates separate duties for providers of general-purpose AI models. Obligations apply in phases between 2025 and 2028.

Illinois AI Safety Measures Act (SB 315)

Illinois Artificial Intelligence Safety Measures Act (SB 315, Public Act 104-0538)

Illinois's frontier-AI safety and transparency law, built on the model of California SB 53: large frontier developers must publish a frontier AI framework and update it every year, publish a transparency report when they deploy a new or substantially modified frontier model, report critical safety incidents to the Illinois Emergency Management Agency and Office of Homeland Security, and protect whistleblowing employees. Unlike the California and New York laws, it also requires annual independent third-party audits, with a published summary.

New York RAISE Act (frontier model safety)

New York Responsible AI Safety and Education (RAISE) Act (S.6953-B / A.6453-B, signed December 2025)

The RAISE Act requires large developers of frontier AI models (defined by training compute and revenue thresholds) to publish and follow a safety and security protocol, report critical safety incidents to the state within a set period, not deploy models that create unreasonable risk of critical harm, and submit to Attorney General enforcement with civil penalties; the chapter amendments create a state oversight office and align definitions with California's Transparency in Frontier AI Act.

Australian mandatory guardrails proposal

Proposals paper: introducing mandatory guardrails for AI in high-risk settings (Australia)

The proposals paper sought views on ten mandatory guardrails for developers and deployers of AI in high-risk settings, a principles-based definition of high-risk AI (with general-purpose AI treated as high-risk), and three regulatory options: adapting existing laws, framework legislation, or a new cross-economy AI Act. The guardrails mirror the Voluntary AI Safety Standard, covering accountability, risk management, data governance, testing, human control, transparency, contestability, supply-chain transparency, records, and conformity assessment.

Beijing general AI measures (2023)

Beijing: Several Measures for Promoting the Innovation and Development of General Artificial Intelligence

北京市促进通用人工智能创新发展的若干措施

Twenty-one measures on compute, data, models, application scenarios and an "inclusive and prudent" approach to regulation and ethics for general-purpose AI in Beijing.

Bletchley Declaration (AI Safety Summit 2023)

The Bletchley Declaration by countries attending the AI Safety Summit, 1–2 November 2023

Signed by countries including the United States, China, the EU, India, Japan and Brazil, the declaration recognises the opportunities of AI and the risks of frontier models, commits signatories to cooperate on identifying AI safety risks and building risk-based policies, and launched the summit series continued in Seoul (2024) and Paris (2025) and the network of AI safety institutes.

Expanded ASEAN Guide on AI Governance and Ethics – Generative AI (2025)

Expanded ASEAN Guide on AI Governance and Ethics – Generative AI

Endorsed in January 2025, the expanded guide addresses risks specific to generative AI (hallucination, deepfakes and misinformation, intellectual property, privacy, security, bias, embedded values) and proposes governance dimensions covering accountability, data, development and deployment, incident reporting, testing and assurance, security, content provenance, safety research and public-interest use, with policy recommendations for member states.

G7 Hiroshima AI Process code of conduct

Hiroshima Process International Code of Conduct for Organizations Developing Advanced AI Systems (G7, 30 October 2023)

Eleven voluntary actions for organisations developing the most advanced AI systems, including foundation and generative models: risk identification and mitigation across the lifecycle, post-deployment vulnerability and incident monitoring, public transparency reports, responsible information sharing, security controls, content authentication such as watermarking, research on societal risks, priority to global challenges, technical standards, and data-input and personal-data safeguards. It sits alongside the Hiroshima Process guiding principles and, from 2025, an OECD-run voluntary reporting framework.

NIST AI RMF

NIST AI Risk Management Framework (AI RMF 1.0) and Generative AI Profile

The NIST AI Risk Management Framework is a voluntary framework for managing risks to individuals, organisations and society from AI. It organises practices into four functions: Govern, Map, Measure and Manage, and describes characteristics of trustworthy AI such as validity, safety, security, accountability, explainability, privacy and fairness. The July 2024 Generative AI Profile (NIST AI 600-1) adds risks and suggested actions specific to generative AI. Many US procurement requirements and state laws reference the framework as a recognised approach.

Singapore Model AI Governance Framework

Model AI Governance Framework (Second Edition) and Model AI Governance Framework for Generative AI

Singapore's Model AI Governance Framework is a voluntary, sector-agnostic guide for organisations deploying AI. The second edition (January 2020) covers four areas: internal governance structures and measures, determining the level of human involvement in AI-augmented decision-making, operations management (data, model development, monitoring), and stakeholder interaction and communication. The May 2024 Model AI Governance Framework for Generative AI extends it with nine dimensions including accountability, data, trusted development and deployment, incident reporting, testing and assurance, security, content provenance, safety and alignment research, and AI for the public good.