# Artificial Intelligence Regulations License: register data (register.json, llms-full.txt) is CC BY 4.0, reuse with a link to artificialintelligenceregulations.com; written entries may be quoted with attribution. AI policy and access terms: https://artificialintelligenceregulations.com/ai-policy.html (see also https://artificialintelligenceregulations.com/ai.txt) CSV export: https://artificialintelligenceregulations.com/data/register.csv German edition (translated from the English version of record, editor-reviewed): https://artificialintelligenceregulations.com/de/index.html, plus /de/about.html, /de/contact.html, /de/privacy.html, /de/ai-policy.html Full machine-readable register: https://artificialintelligenceregulations.com/llms-full.txt RSS: https://artificialintelligenceregulations.com/feed.xml > A free, continuously maintained repository of artificial intelligence laws and regulations worldwide. Every entry is categorized by jurisdiction and topic, cites the governing statute, and links to the primary legal text. Free to use. Last updated September 14, 2026. Cited to primary legal sources; not legal advice. ## Global picture (as of August 2026) - Three comprehensive AI laws are in force: the EU AI Act (Regulation (EU) 2024/1689), South Korea's AI Basic Act (effective 22 January 2026), and Vietnam's AI Law (effective 1 March 2026, the first in Southeast Asia, with extraterritorial reach). - Taiwan's AI Basic Act (in force 14 January 2026) and Peru's Law 31814 with its January 2026 regulation are framework statutes: binding principles with obligations phasing in. - China enforces the most extensive set of binding sectoral AI rules but has no single comprehensive statute. - The United States has no comprehensive federal AI law; it regulates mainly through state law (11 states tracked, led by California, Texas, Illinois, and New York) plus executive orders, including EO 14365 (December 2025) seeking to preempt state laws. Virginia's comprehensive bill was vetoed; Colorado repealed its act before effect. - The United Kingdom, Japan, Israel, Switzerland, and Australia use sectoral or principles-based approaches with no horizontal AI act; Australia formally shelved mandatory guardrails in its December 2025 National AI Plan. - The EU deferred high-risk (Annex III) obligations to 2 December 2027 via the 2026 Digital Omnibus; Article 50 transparency duties apply from 2 August 2026; the Commission adopted the final Article 50 guidelines on 20 July 2026 and assessed the Code of Practice on Transparency as adequate. - This register covers 33 national and supranational regimes plus 11 US states, each cited to primary sources. ## Jurisdictions - [China](https://artificialintelligenceregulations.com/jurisdictions/china.html): In force. Extensive binding sectoral rules; no single comprehensive act. - [European Union](https://artificialintelligenceregulations.com/jurisdictions/european-union.html): Phasing in. Comprehensive, horizontal, risk-based statute. - [South Korea](https://artificialintelligenceregulations.com/jurisdictions/south-korea.html): In force. Comprehensive risk-based act with a light enforcement posture. - [United Kingdom](https://artificialintelligenceregulations.com/jurisdictions/united-kingdom.html): No comprehensive law. Sectoral, principles-based; no horizontal AI act by design. - [United States](https://artificialintelligenceregulations.com/jurisdictions/united-states-federal.html): No comprehensive law. No comprehensive law; deregulatory posture and state preemption push. - [Australia](https://artificialintelligenceregulations.com/jurisdictions/australia.html): No comprehensive law. No AI act; existing laws plus voluntary guidance, by explicit choice. - [Brazil](https://artificialintelligenceregulations.com/jurisdictions/brazil.html): Proposed. Comprehensive EU-style bill advancing through the legislature. - [Canada](https://artificialintelligenceregulations.com/jurisdictions/canada.html): Lapsed / failed. Proposed federal act lapsed; voluntary code plus provincial rules. - [Council of Europe](https://artificialintelligenceregulations.com/jurisdictions/council-of-europe.html): Partially in force. First binding international AI treaty (framework convention). - [Japan](https://artificialintelligenceregulations.com/jurisdictions/japan.html): In force. Innovation-first framework law; light-touch and non-punitive. - [Mexico](https://artificialintelligenceregulations.com/jurisdictions/mexico.html): Proposed. No AI law; bills pending, data protection reform added ADM rights. - [Peru](https://artificialintelligenceregulations.com/jurisdictions/peru.html): In force. First AI law in Latin America; promotional statute with risk-based regulation. - [Saudi Arabia](https://artificialintelligenceregulations.com/jurisdictions/saudi-arabia.html): No comprehensive law. Centralized governance through SDAIA; principles and drafts, no AI act. - [Switzerland](https://artificialintelligenceregulations.com/jurisdictions/switzerland.html): Proposed. Deliberate counter-model to the EU: sectoral, with a treaty-implementation bill due. - [Taiwan](https://artificialintelligenceregulations.com/jurisdictions/taiwan.html): In force. Framework statute; principles in force, obligations to follow. - [UAE](https://artificialintelligenceregulations.com/jurisdictions/united-arab-emirates.html): Partially in force. Layered model: charters and strategy federally, binding rules in zones and finance. - [Vietnam](https://artificialintelligenceregulations.com/jurisdictions/vietnam.html): In force. Comprehensive, risk-based AI law with extraterritorial reach. - [Argentina](https://artificialintelligenceregulations.com/jurisdictions/argentina.html): No comprehensive law. No AI law; guidance and hearings, momentum without a vehicle. - [Chile](https://artificialintelligenceregulations.com/jurisdictions/chile.html): Proposed. EU-inspired risk-based bill in Congress; updated national policy. - [Colombia](https://artificialintelligenceregulations.com/jurisdictions/colombia.html): Proposed. Three bills in play; policy framework set, no law. - [Egypt](https://artificialintelligenceregulations.com/jurisdictions/egypt.html): No comprehensive law. Strategy and charter; data protection law binds, no AI act. - [Hong Kong](https://artificialintelligenceregulations.com/jurisdictions/hong-kong.html): No comprehensive law. No AI law; dense sectoral guidance, hardest in finance. - [India](https://artificialintelligenceregulations.com/jurisdictions/india.html): No comprehensive law. Principles-based guidelines plus targeted synthetic-media rules. - [Indonesia](https://artificialintelligenceregulations.com/jurisdictions/indonesia.html): Proposed. Roadmap and ethics guidance; binding rules in development. - [Israel](https://artificialintelligenceregulations.com/jurisdictions/israel.html): No comprehensive law. Deliberately decentralized: sectoral regulators, soft law, no AI act. - [Kenya](https://artificialintelligenceregulations.com/jurisdictions/kenya.html): Proposed. AI Bill tabled 2026: classification, high-risk duties, and a register. - [Malaysia](https://artificialintelligenceregulations.com/jurisdictions/malaysia.html): No comprehensive law. Voluntary governance guidelines; no binding AI law. - [New Zealand](https://artificialintelligenceregulations.com/jurisdictions/new-zealand.html): No comprehensive law. No AI-specific law; existing technology-neutral statutes. - [Nigeria](https://artificialintelligenceregulations.com/jurisdictions/nigeria.html): Proposed. Binding framework in motion: risk-based bill with licensing for high-risk AI. - [Singapore](https://artificialintelligenceregulations.com/jurisdictions/singapore.html): No comprehensive law. Voluntary frameworks and testing tools; no binding AI law. - [South Africa](https://artificialintelligenceregulations.com/jurisdictions/south-africa.html): No comprehensive law. Draft national AI policy; POPIA binds, no AI legislation proposed yet. - [Thailand](https://artificialintelligenceregulations.com/jurisdictions/thailand.html): Proposed. Draft AI law in development; data protection already applies. - [Turkey](https://artificialintelligenceregulations.com/jurisdictions/turkey.html): Proposed. Draft AI framework pending; data protection and cybersecurity law bind now. ## Topics - [Prohibited AI practices](https://artificialintelligenceregulations.com/topics/prohibited-practices.html): Uses of AI that are banned outright rather than merely regulated. - [High-risk systems](https://artificialintelligenceregulations.com/topics/high-risk-systems.html): AI used in consequential decisions: hiring, credit, health, and public services. - [Transparency and disclosure](https://artificialintelligenceregulations.com/topics/transparency-disclosure.html): Telling people when they are dealing with AI or AI-generated content. - [Deepfakes and synthetic media](https://artificialintelligenceregulations.com/topics/deepfakes-synthetic-media.html): Rules on AI-generated likenesses, voices, and non-consensual imagery. - [Biometrics and facial recognition](https://artificialintelligenceregulations.com/topics/biometrics-facial-recognition.html): Facial recognition, emotion recognition, and biometric categorization. - [Automated decisions and employment](https://artificialintelligenceregulations.com/topics/automated-decisions-employment.html): AI in hiring, performance, promotion, and termination. - [Generative AI and foundation models](https://artificialintelligenceregulations.com/topics/generative-foundation-models.html): Rules aimed at general-purpose models and the companies that build them. ## United States - [US state AI laws (state-by-state tracker)](https://artificialintelligenceregulations.com/jurisdictions/us-states.html): 11 states with status, effective dates, coverage, and enforcement: California, Colorado, Connecticut, Illinois, Montana, New York, Tennessee, Texas, Utah, Virginia, Washington. - US states are tracked separately from the federal picture. Most state laws follow the affected resident, not the company's location. ## Reference - [Deadline calendar](https://artificialintelligenceregulations.com/deadlines.html): every AI-law effective date worldwide, February 2025 through August 2028. - [Penalty comparison](https://artificialintelligenceregulations.com/penalties.html): fines and enforcement compared across all tracked jurisdictions. - [September 2026 briefing](https://artificialintelligenceregulations.com/briefing-september-2026.html): the monthly review, what changed, what was corrected, what comes next. - [EU vs US](https://artificialintelligenceregulations.com/eu-ai-act-vs-us.html): the EU AI Act and US regulation compared on scope, enforcement, penalties, reach, and dates. - [US questions](https://artificialintelligenceregulations.com/us-ai-law-questions.html): 12 question-shaped answers on American AI law, from the federal gap to BIPA to the frontier template. - [Law index](https://artificialintelligenceregulations.com/laws.html): every named instrument on the register, 78 laws across 44 jurisdictions, each linked to its entry. - [Glossary](https://artificialintelligenceregulations.com/glossary.html): 53 terms of art defined as the statutes use them, each linked to the governing law. ## Structured data - [Machine-readable register (JSON)](https://artificialintelligenceregulations.com/data/register.json): every jurisdiction and law with status, effective dates, penalties, and primary-source links. Updated with the site. - Every page carries Schema.org JSON-LD: Legislation entities with legal-force status, FAQPage, BreadcrumbList, ItemList, DefinedTermSet, and Dataset markup. ## Other pages - [News digest](https://artificialintelligenceregulations.com/news.html): curated summaries of recent AI regulation developments, plus a live feed that refreshes every six hours. - [About and methodology](https://artificialintelligenceregulations.com/about.html): how entries are built, sourced, and dated. ## Publisher - Maintained by the team behind Agency Echelon (https://agencyechelon.com), a digital consultancy specializing in search and AI visibility.