AI regulation by country: EU, UK, US and Canada compared
Four big markets regulate AI in four different ways: a comprehensive EU statute, a UK principles-and-regulators model, a US mix of an executive order and state laws, and a Canadian federal law that did not pass. This page compares them using primary sources and says plainly what we could not confirm. Details on the EU are in the EU AI Act explained and its timeline; the topic hub is AI law, copyright and safety.
By the Chelure Enterprises Editorial Team. Last reviewed . How we check our sources.
At a glance
| Jurisdiction | Approach | Binding AI-specific law? | Where it stands (checked {UPDATED_HUMAN}) |
|---|---|---|---|
| European Union | One comprehensive statute with risk levels | Yes: Regulation (EU) 2024/1689, amended by Regulation (EU) 2026/1744 | In force since 1 August 2024; high-risk rules for Annex III now apply from 2 December 2027 |
| United Kingdom | Non-statutory principles applied by existing regulators | No dedicated statute found | 2023 white paper and 2024 response; one Private Member's Bill with 'artificial intelligence' in its title in the 2026-27 session |
| United States (federal) | Executive policy plus pressure on state laws | No federal AI statute confirmed in our sources; Executive Order 14365 | EO signed 11 December 2025; legislative proposal for a uniform framework requested |
| United States (states) | Individual state laws | Yes, state by state (examples below) | California SB 53 chaptered 29 September 2025; Colorado SB26-189 signed 14 May 2026 |
| Canada (federal) | No federal AI statute after AIDA | No AI-specific statute found | AIDA, in Bill C-27, did not pass; privacy bill C-36 listed on LEGISinfo |
European Union
The AI Act sets four risk levels, bans specific practices, regulates high-risk systems and sets transparency duties, with fines up to EUR 35 million or 7 percent of worldwide annual turnover. It applies to providers placing AI systems on the EU market whatever their location, and to third-country providers and deployers where the output is used in the Union. The 2026 Digital Omnibus delayed the high-risk dates. See our explainer for the full picture.
United Kingdom
The UK government's March 2023 white paper, A pro-innovation approach to AI regulation, set out five cross-sector principles: safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress. They were issued on a non-statutory basis for existing regulators to implement, not put on a statutory footing initially. The government's February 2024 response kept the non-statutory approach and was followed by guidance for regulators on applying the principles.
We checked Parliament's bills website on the date of this page by searching bill titles for "artificial intelligence". It returned one bill in the current 2026-27 session: the Artificial Intelligence (Consent and Misuse) Bill, a Private Member's Bill that originated in the House of Commons and awaits second reading. A title search cannot show every law that touches AI, so this is not proof that nothing else applies; confirm the current position on gov.uk and Parliament's site.
United States
Federal. Executive Order 14365, Ensuring a National Policy Framework for Artificial Intelligence, signed December 11, 2025, states a policy of a minimally burdensome national framework and directs several actions: an Attorney General task force to challenge state AI laws (within 30 days), a Commerce evaluation identifying onerous state laws (within 90 days), conditions on certain broadband funding for states with onerous AI laws, an FCC proceeding on a federal reporting and disclosure standard, an FTC policy statement on when state laws that require altering truthful AI outputs are preempted, and a legislative recommendation for a uniform federal framework. The recommendation should not preempt otherwise lawful state laws on child safety, AI compute and data centre infrastructure (apart from general permitting), or state government use of AI. An executive order is not a statute, so state laws continue to apply as enacted.
California. SB 53, the Transparency in Frontier Artificial Intelligence Act, was chaptered on September 29, 2025. It covers frontier models trained with more than 10^26 operations (counting later fine-tuning); large frontier developers, those with annual gross revenues above $500 million, must write and publish a frontier AI framework; frontier developers must report critical safety incidents to the Office of Emergency Services within 15 days of discovery (24 hours if there is an imminent risk of death or serious physical injury); and civil penalties of up to $1,000,000 per violation can be recovered only in an action brought by the Attorney General.
Colorado. SB26-189, signed May 14, 2026, repeals and reenacts the 2024 AI consumer-protection law (SB24-205) as a law on automated decision-making technology used in consequential decisions such as employment, housing, lending, insurance, health care and education. From January 1, 2027, developers must give deployers technical documentation, deployers must give consumers notice at the point of interaction, and consumers can request correction of incorrect data and human review after an adverse outcome. The Attorney General enforces it as a deceptive trade practice, with a 60-day notice-and-cure period for actions filed before January 1, 2030, and it creates no new private right of action.
Canada
Canada's proposed Artificial Intelligence and Data Act (AIDA) was one of three acts in Bill C-27. A government consultation report states that parliamentary work on AIDA ceased with prorogation, and Bill C-27 was among the government bills before the House when Parliament was prorogued in January 2025. Parliament's LEGISinfo lists Bill C-36, the Protecting Privacy and Consumer Data Act, sponsored by the Minister of Artificial Intelligence and Digital Innovation; the page does not say whether it contains AI or automated-decision provisions, we have not read the bill text, and its dates on that page are inconsistent, so check LEGISinfo directly.
What this means if you build or sell AI products
- Map your markets first. The EU AI Act reaches providers outside the EU, so a US or Canadian company selling into Europe can be covered.
- Track US state laws individually. California and Colorado differ in scope: frontier-model transparency in one, automated consequential decisions in the other.
- Use a voluntary framework as a common base. See the NIST AI Risk Management Framework.
- Re-check dates. Several of these positions changed during 2026.
This is a plain-language comparison, not legal advice. Laws change; the status of each item was checked on November 13, 2026 against the sources below.
Frequently asked questions
Which country has the strictest AI law?
The EU has the most comprehensive binding AI statute we can confirm: the AI Act, Regulation (EU) 2024/1689, which sorts AI into risk levels and sets fines up to EUR 35 million or 7 percent of worldwide turnover. The UK, US federal government and Canada have no equivalent single statute in the sources we checked.
Does the UK have an AI law?
Not a dedicated AI statute that we found. The government's 2023 white paper set out five non-statutory principles for existing regulators to apply, and a title search of Parliament's bills site on the date of this page returned one bill with 'artificial intelligence' in its title in the current session, a Private Member's Bill, the Artificial Intelligence (Consent and Misuse) Bill. Other laws can still apply to AI without naming it.
Is there a US federal AI law?
The federal action we can confirm from primary sources is an executive order, EO 14365 of December 11, 2025, which directs agencies to challenge and discourage conflicting state AI laws and asks for a legislative proposal for a uniform federal framework. An executive order is not a statute. State laws continue to apply as enacted.
What happened to Canada's AI law?
The Artificial Intelligence and Data Act (AIDA) was part of Bill C-27, which did not pass; a government consultation report says parliamentary work on AIDA ceased with prorogation. Parliament's LEGISinfo lists a separate privacy bill, C-36, the Protecting Privacy and Consumer Data Act, which we have not read for AI provisions.
Which rules apply to a small developer selling worldwide?
Possibly several. The EU AI Act applies to providers placing systems on the EU market wherever they are established, and to third-country providers whose output is used in the Union. US state laws apply to what they cover in their states, and the UK relies on existing regulators. Check each market you sell into.
Sources
- European Commission: AI Act
- White House: Executive Order 14365, Ensuring a National Policy Framework for Artificial Intelligence
- California Legislative Information: SB 53 bill text
- Colorado General Assembly: SB26-189 Automated Decision-Making Technology
- GOV.UK: A pro-innovation approach to AI regulation (white paper)
- GOV.UK: government response to the consultation (2024)
- UK Parliament: bills search for 'artificial intelligence'
- Parliament of Canada: LEGISinfo, Bill C-36
- Innovation, Science and Economic Development Canada: consultation report mentioning AIDA
Primary sources opened on November 13, 2026. Prices come from our verified price table (see the methodology).