AI regulations around the world: country guide (2026)

The short answer: the European Union is the only major jurisdiction phasing in a comprehensive, binding AI law, while South Korea and Japan have passed framework acts. The US has no federal AI statute, so states are writing the rules. China regulates through a series of targeted measures rather than one law. The UK, India, Canada and Türkiye have no AI-specific law in force, and Brazil's bill has cleared the Senate but not the lower house.

A caveat up front: this overview is based mainly on developments through late 2025 and could not be verified against current sources as of 14 October 2026. Many 2026 dates below are scheduled dates, not confirmed outcomes. Check official sources before relying on any of them for compliance.

When does the EU AI Act apply?

The EU AI Act (Regulation (EU) 2024/1689) was published on 12 July 2024 and entered into force on 1 August 2024. It sorts AI by risk: some practices are banned outright, high-risk systems face strict requirements, and certain systems carry transparency duties.

Obligations arrive in stages. Bans on unacceptable-risk practices and AI literacy duties started on 2 February 2025. Rules for general-purpose AI (GPAI) models, plus governance and penalties, followed on 2 August 2025. Most remaining obligations, including Annex III high-risk systems and Article 50 transparency, were set for 2 August 2026, with high-risk AI embedded in regulated products due on 2 August 2027.

Penalties reach €35 million or 7% of global annual turnover for banned practices, €15 million or 3% for most other breaches, and €7.5 million or 1% for supplying incorrect information. A voluntary GPAI Code of Practice was published on 10 July 2025; most large labs signed, Meta declined, and xAI signed only the safety chapter.

The big open question is the Digital Omnibus, proposed by the Commission on 19 November 2025. It would tie high-risk deadlines to the availability of standards, with long-stop dates of roughly December 2027 for Annex III and August 2028 for Annex I. We could not confirm whether Parliament and Council adopted it, so verify whether the August 2026 obligations actually took effect.

Does the US have a federal AI law?

No. Federal AI policy runs mostly through executive orders. President Trump revoked Biden's EO 14110 on 20 January 2025 and signed EO 14179 on removing barriers to AI three days later. The AI Action Plan arrived on 23 July 2025 alongside three orders, including one targeting "woke AI" in federal procurement.

A proposed 10-year moratorium on state AI laws was stripped from the budget bill after a 99–1 Senate vote on 1 July 2025. An executive order signed in December 2025 set out to challenge state laws, including through a Justice Department litigation task force and conditions on broadband funding; whether lawsuits or congressional preemption followed is unclear. The one notable federal statute is the TAKE IT DOWN Act, signed 19 May 2025, which criminalises non-consensual intimate imagery including deepfakes and requires platforms to remove it within 48 hours, a duty applying from around May 2026.

Which US states regulate AI?

California has moved furthest. SB 53, the frontier AI transparency law, and AB 2013 on training-data disclosure took effect on 1 January 2026, and SB 243 on companion chatbots was signed in October 2025. Its privacy agency also finalised automated decision-making rules in 2025, phased in through 2027.

Texas's TRAIGA and Illinois's HB 3773 on AI in hiring both took effect on 1 January 2026. Colorado's AI Act was delayed in August 2025 to 30 June 2026; we could not confirm whether it was changed again. New York's RAISE Act was signed with amendments in December 2025, and Utah amended its 2024 AI Policy Act in 2025.

How does China regulate AI?

China has no single AI law; a comprehensive one has appeared on the State Council's legislative plan but, as far as we know, has not been enacted. Instead it uses targeted rules: algorithmic recommendation provisions (March 2022), deep synthesis provisions (January 2023) and the Interim Measures for Generative AI (15 August 2023), which require services to be filed with regulators.

Labeling measures for AI-generated content took effect on 1 September 2025, backed by the mandatory national standard GB 45438-2025. Cybersecurity Law amendments adding AI provisions were reported to take effect on 1 January 2026; that should be confirmed.

What are the rules in the UK, Japan and South Korea?

The UK takes a principles-based, "pro-innovation" approach with no AI statute, leaving enforcement to existing regulators. It published an AI Opportunities Action Plan in January 2025 and renamed its AI Safety Institute the AI Security Institute in February 2025. A promised frontier AI bill has been repeatedly delayed, and the AI and copyright question remained unresolved at last check.

Japan passed its AI Promotion Act on 28 May 2025, fully in force from September 2025. It is non-binding, carries no penalties and created an AI Strategy Headquarters chaired by the Prime Minister. An AI Basic Plan was expected around December 2025.

South Korea's AI Basic Act passed on 26 December 2024 and was scheduled to take effect on 22 January 2026. It covers "high-impact" AI and labeling of generative output, and requires large foreign providers to appoint a local representative. The maximum fine is ₩30 million (about $20,000), and the government announced a grace period of at least a year on fines.

Where do India, Brazil, Canada and Türkiye stand?

India has no AI-specific law. Its IT ministry published principles-based AI Governance Guidelines on 5 November 2025, and an October 2025 draft rule would require visible labels covering 10% of synthetic images or audio; whether it was finalised is unclear. Data protection rules were notified in November 2025 with 12–18 months to comply.

Brazil's risk-based bill PL 2338/2023, modelled on the EU, passed the Senate on 10 December 2024 and went to a special committee in the Chamber of Deputies. October 2026 elections may slow it further. In Canada, the AIDA bill died when Parliament was prorogued on 6 January 2025; AI minister Evan Solomon, appointed in May 2025, has signalled light regulation, leaving a 2023 voluntary generative AI code in place.

Türkiye has no AI law in force. Policy rests on the National AI Strategy 2021–2025 and its 2024–2025 Action Plan. A short, EU-inspired bill filed in parliament in June 2024, with fines up to ₺35 million or 7% of turnover, did not advance, and 2025 proposals on deepfakes and criminal liability need status checks. At last check Türkiye had not signed the Council of Europe AI Convention.

Internationally, that Convention opened for signature on 5 September 2024 and was signed by the EU, UK, Canada, Japan and, under Biden, the US, whose current stance is unclear. The G7 Hiroshima code, the OECD AI Principles and UN bodies created in August 2025 offer guidance, but binding obligations still come from national law.