Global AI Regulation: Simplified and Tracked
Follow AI regulation as it happens, in 40+ countries and counting.
The map identifies only those jurisdictions where there is evidence that either: a draft law targeting AI has been prepared (in yellow); or a draft law is under consideration by the relevant government/parliamentary authorities (blue); or legislation has been enacted (green).
For those countries where none of the above actions apply, they remain in grey.
The map is updated regularly in line with the detailed country-by-country analysis below.
Global Regulation Status Tracker (Beta)
Version 3 (21 January 2025)
Notes: We are working to include all significant international agreements for AI cooperation and/or governance, such as the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, or the African Union Continental Artificial Intelligence Strategy.
| Jurisdiction | Regulation Status | Summary | Analysis |
|---|---|---|---|
| Austria | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Austria needs to implement several key measures, including the following important steps as part of Austria’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Bahrain | Draft bill proposed | National legislation: 28 April 2024 – AI Regulation law approved by the Shura Council. | |
|
National Legislation:
28 April 2024 – The AI Regulation law focuses on regulating misuse, including the introduction of penalties and sentencing for various activities including operating or programming AI systems without a licence or misuses such as discrimination. (https://www.zawya.com/en/world/middle-east/new-law-to-regulate-use-of-ai-approved-in-bahrain-djpvmm4r) |
|||
| Brazil | In progress | International treaty/declaration: 9 August 2024 – ‘Cartagena de Indias Declaration’. National legislation: 3 May 2023 – Draft bill No. 2338, of 2023. |
|
|
International treaty/declaration:
9 August 2024 – ‘Cartagena de Indias Declaration for Governance, the construction of AI ecosystems and the promotion of AI education in an Ethical and Responsible manner in Latin America and the Caribbean’ (the “Declaration”); signed by 17 countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Curazao, Ecuador, Guatemala, Guyana, Honduras, Panama, Paraguay, Peru, Dominican Republic, Suriname and Uruguay. The Declaration is a framework which represents Latin America’s proactive stance in addressing AI challenges while promoting responsible innovation. It emphasises the importance of digital skills development and knowledge sharing across borders. Core Components:
National Legislation:
3 May 2023 – Draft bill No. 2338, of 2023 establishes a comprehensive framework for responsible AI development and use, emphasising protection of fundamental rights while promoting innovation. Key features include: Core Principles:
The bill introduces a risk-based regulatory framework categorising AI systems into:
|
|||
| Bulgaria | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Bulgaria needs to implement several key measures, including the following important steps as part of Bulgaria’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Canada | In progress | National legislation: 16 June 2023 – Bill C-27 for the draft Artificial Intelligence and Data Act (“AIDA”) proposed. | |
|
National Legislation:
16 June 2023 – Bill C-27, the draft Artificial Intelligence and Data Act (“AIDA”) is currently before the House of Commons (https://parl.ca/legisinfo/en/bill/44-1/c-27 ) If passed, it is anticipated that the law would be implemented by 2025. The current draft text of AIDA outlines a comprehensive framework for AI regulation, focusing on high-impact systems that could affect human rights, health, and safety. The legislation will regulate interprovincial and international AI trade, requiring mandatory assessments, risk mitigation, and transparency measures. Key oversight targets include screening systems, biometric identification, behavioural influence systems, and critical health/safety applications. Implementation is expected by 2025, with an AI Data Commissioner overseeing compliance and enforcement through administrative penalties. AIDA is broken down into three parts: • Part 1 enacts the Consumer Privacy Protection Act to govern the protection of personal information of individuals; • Part 2 enacts the Personal Information and Data Protection Tribunal Act, to establish an administrative tribunal to hear appeals of certain decisions made by the Privacy Commissioner under the Consumer Privacy Protection Act and to impose penalties for the contravention of certain provisions of that Act; • Part 3 enacts the Artificial Intelligence and Data Act to regulate international and interprovincial trade and commerce in AI systems, requiring certain measures to mitigate risks of harm and biased output related to “high-impact artificial intelligence systems”. The legislation also includes provisions relating to public reporting including the power for the Minister to order production of records related to AI systems. That Act also establishes prohibitions related to the possession or use of illegally obtained personal information for the purpose of designing, developing, using or making available for use an AI system which causes serious harm to individuals https://www.parl.ca/DocumentViewer/en/44-1/bill/C-27/first-reading |
|||
| Chile | In progress | International treaty/declaration: 9 August 2024 – ‘Cartagena de Indias Declaration for Governance, the construction of AI ecosystems and the promotion of AI education in an Ethical and Responsible manner in Latin America and the Caribbean’; signed by Chile (alongside 16 other Latin American countries).
National legislation: 7 May 2024 – Draft bill regulating Artificial Intelligence systems (Nr. 16821-19) introduced to the Chilean Chamber of Deputies |
|
|
International treaty/declaration:
9 August 2024 – ‘Cartagena de Indias Declaration for Governance, the construction of AI ecosystems and the promotion of AI education in an Ethical and Responsible manner in Latin America and the Caribbean’ (the “Declaration”); signed by 17 Latin American countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Curazao, Ecuador, Guatemala, Guyana, Honduras, Panama, Paraguay, Peru, Dominican Republic, Suriname and Uruguay. The Declaration is a framework which represents Latin America’s proactive stance in addressing AI challenges while promoting responsible innovation. It emphasises the importance of digital skills development and knowledge sharing across borders. Core Components:
https://mintic.gov.co/portal/715/articles-383990_recurso_1.pdf
National Legislation:
7 May 2024 – The draft bill regulating Artificial Intelligence systems (Nr. 16821-19) including cybersecurity requirements introduces comprehensive AI regulation that aims to promote responsible AI development while protecting fundamental rights. The legislation establishes a risk-based framework (similar to the EU) categorising AI systems into four levels:
Key features include:
The bill emphasises core principles including:
The legislation would take effect the first business day of the year following its publication, with implementation regulations to be developed within 12 months. The bill is currently under consideration. https://dig.watch/updates/chile-introduces-updated-national-ai-policy-and-new-ai-legislation |
|||
| China | Enacted | National legislation: 31 December 2021 – Provisions on the Administration of Algorithm-generated Recommendations for internet Information Services issued (effective 1 March 2022). 25 November 2022 – Internet Information Service Deep Synthesis Management Provisions 2022 issued (effective 1 January 2023). 10 July 2023 – Interim Measures for the Management of Generative Artificial Intelligence Services 2023 (effective 15 August 2023). |
|
|
National Legislation:
31 December 2021 – The Provisions on the Administration of Algorithm-generated Recommendations for internet Information Services, which came into effect on 1 March 2022, regulate algorithmic decision making via the use of algorithmic recommendation services within online content, such as those used by messaging apps or on e-commerce sites in mainland China. Among the provisions is the requirement for algorithms used to generate recommendations to be registered with the Cyberspace Administration of China (CAC). As a result of the regulations, users of online content are given a choice as to whether to opt-in or out of algorithmic recommendation services. Users may also view and delete the keywords used to profile them as individuals. The regulations also target algorithms which may be used in search-filtering and personal recommendations, and those that generate content including ‘deepfakes’, which under the regulations must be labelled as algorithmically generated. The regulations also seek to prevent ‘fake news’, content aimed at causing controversy, or addictive content for children (https://technode.com/2022/04/05/digital-policy-experts-weigh-in-on-chinas-new-algorithm-regulation/ (https://library.fes.de/pdf-files/bueros/bruessel/19904.pdf ) 25 November 2022 – Internet Information Service Deep Synthesis Management Provisions 2022 came into effect on 1 January 2023. The provisions establish a comprehensive regulatory framework for technologies utilising generative sequencing algorithms which may be used to create ‘deep fakes’. The scope encompasses text generation, speech conversion, biometric manipulation, image generation, and digital character creation. Service providers and technical supporters must adhere to strict protocols regarding data security, personal information protection, and content management. Key obligations include obtaining specific consent for biometric information editing, implementing robust technical management systems, and conducting security assessments for functions that may impact national security or public interests. Providers of technology which may be in a position to influence public opinion are subject to additional registration requirements under the algorithmic recommendation provisions (see above and below) (https://www.loc.gov/item/global-legal-monitor/2023-04-25/china-provisions-on-deep-synthesis-technology-enter-into-effect/ ). 10 July 2023 – The Measures for the Management of Generative Artificial Intelligence Services 2023 came into effect on 15 August 2023. Although they are labelled as “interim”, they remain in effect. These regulations aim to promote controlled development of generative AI while safeguarding national security and public interests. The measures specifically target services generating text, image, audio, and video content for public consumption within mainland China, explicitly excluding private development and applications. The provisions provide inter alia that providers with “public opinion attributes or social mobilisation capabilities” must undergo CAC (Cyberspace Administration of China) security assessment and comply with Algorithmic Recommendation Provisions. The measures establish provisions for addressing non-compliant services originating outside mainland China through technical interventions. The regulation includes provisions that address generative AI services provided to the public within mainland China but which originate from outside the territory of mainland China. Should such services contravene any Chinese law, regulation or generative AI measures, the regulations provide that the national cyberspace authority shall “notify the relevant bodies to employ technical measures and other necessary measures to deal with the matter.” (Art. 20.) (https://www.loc.gov/item/global-legal-monitor/2023-07-18/china-generative-ai-measures-finalized/ ) |
|||
| Croatia | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Croatia needs to implement several key measures, including the following important steps as part of Croatia’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Cyprus | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Cyprus needs to implement several key measures, including the following important steps as part of Cyprus’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Czech Republic | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, the Czech Republic needs to implement several key measures, including the following important steps as part of the Czech Republic’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Denmark | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Denmark needs to implement several key measures, including the following important steps as part of Denmark’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Estonia | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Estonia needs to implement several key measures, including the following important steps as part of Estonia’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| European Union | Enacted | International treaty/declaration: 5 September 2024 – Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (“AI Convention”) concluded. EU Regulation: 1 August 2024 – The AI Act (Regulation (EU) 2024/1689) entered into force on 1 August 2024 and will become generally applicable from 2 August 2025, though certain provisions are subject to early or delayed implementation where explicitly stated. Select “show analysis” for more. |
|
|
International treaty/declaration:
5 September 2024 – Members of the Council of Europe, including EU Member States and the UK (as well as the US) concluded the Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (“AI Convention”). The AI Convention aims to ensure that activities within the lifecycle of AI systems comply with fundamental principles including human rights, democracy and the rule of law.
EU Regulation:
1 August 2024 – The AI Act (Regulation (EU) 2024/1689) became effective. The Act establishes the EU’s comprehensive framework for AI regulation. The Act categorises AI systems based on risk levels:
The legislation aims to foster responsible AI development while reducing business burdens through clear requirements. It promises benefits across sectors:
By implementing this regulatory framework based on human rights and fundamental values, the EU positions itself as a global leader in safe AI development. The timeline for implementation is as follows:
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ%3AL_202401689 The provisions of the AI Act are directly applicable in all EU Member States, though some national legislation may be required to give effect to certain narrow parts of the AI Act e.g Recital 96 and Article 27(10) concerning biometric identification systems. In the meantime, the key responsibilities for Member States include the requirement to:
Each EU Member State’s progress in achieving the above will be detailed within the individual country analysis of this table. |
|||
| Finland | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Finland needs to implement several key measures, including the following important steps as part of Finland’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| France | Enact | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, France needs to implement several key measures, including the following important steps as part of France’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Germany | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Germany needs to implement several key measures, including the following important steps as part of Germany’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Greece | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Greece needs to implement several key measures, including the following important steps as part of Greece’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Hong Kong | None | National legislation: None | |
| Hungary | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Hungary needs to implement several key measures, including the following important steps as part of Hungary’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| India | None | National legislation: None | |
| Ireland | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Ireland needs to implement several key measures, including the following important steps as part of Ireland’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Italy | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Italy needs to implement several key measures, including the following important steps as part of Italy’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Kenya | In progress | International treaty/declaration: 19 July 2024 – Though not a treaty or declaration, it is worth noting that 55 African nations, including Kenya, joined the ‘African Union Continental Strategy on Artificial Intelligence’ endorsed by the African Union Executive Council. National legislation: 4 September 2023 – Draft Robotics and Artificial Intelligence Society Bill 2023 published. |
|
|
International treaty/declaration: 19 July 2024 – The Continental AI Strategy outlines a comprehensive framework for harnessing AI’s potential while addressing its challenges. The strategy aims to serve as a blueprint for AI regulation across the African Union, and focuses on five key areas: 1. Maximising AI benefits for socio-economic development and cultural renaissance 2. Minimising risks and ensuring responsible AI development 3. Building capabilities in infrastructure, datasets, skills, and innovation 4. Fostering regional/international cooperation 5. Accelerating AI investment The strategy emphasises: – People-centric development approach – Protection of human rights and dignity – Inclusion and diversity – Gender equality – Safety and security considerations – Cultural preservation Key Implementation Aspects: – Two-phase implementation (2025-2030) – Development of national AI strategies – Creation of regulatory frameworks – Establishment of AI governance systems – Focus on capacity building in education and workforce – Promotion of research and innovation – Development of AI infrastructure and data ecosystems The strategy aims to position Africa as an integral participant in the global AI landscape while ensuring AI development serves African interests and values (https://au.int/en/pressreleases/20240617/african-ministers-adopt-landmark-continental-artificial-intelligence-strategy ).
National legislation:
4 September 2023 – The draft Robotics and Artificial Intelligence Society Bill 2023 is still under consideration following publication in September 2023. The draft bill provides for the establishment of the Kenya Robotics and Artificial Intelligence Society, to regulate and promote robotics and AI practices in the country. The draft bill aims to set industry standards to ensure responsible development and application of AI technologies. The current text also aims to introduce mandatory registration for practitioners in the robotics and AI sectors, with penalties for unlicensed entities, including fines up to KES 1 million (approximately USD 6,000) or imprisonment for up to two years (https://kenyaai.ke/docs/kenya-robotics-and-artificial-intelligence-society-bill-2023/) (http://www.parliament.go.ke/sites/default/files/2024-02/Hansard%20Report%20-%20Thursday%2C%2030th%20November%202023%20%28P%29.pdf) |
|||
| Latvia | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Latvia needs to implement several key measures, including the following important steps as part of Latvia’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Lithuania | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Lithuania needs to implement several key measures, including the following important steps as part of Lithuania’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Luxembourg | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Luxembourg needs to implement several key measures, including the following important steps as part of Luxembourg’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Malta | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Malta needs to implement several key measures, including the following important steps as part of Malta’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Mexico | Draft bill proposed | National legislation: 27 February 2024 – Draft Federal Law Regulating Artificial Intelligence proposed. | |
|
National Legislation:
27 February 2024 – The draft Federal Law Regulating Artificial Intelligence provides the following: “a) …extraterritorial effect, so the law would be applicable even to providers of Artificial Intelligence Systems (AIS) located abroad that offer services in Mexico or whose information generated is used in Mexico. b) The regulator in charge of authorising AIS providers will be the Federal Telecommunications Institute (IFT, in its Spanish acronym). In addition, the bill proposes the creation of a National Artificial Intelligence Commission, as an advisory body to the IFT, which will be formed by scientists. c) The bill proposes to classify AIS according to the risks they may trigger, similar to the European Union (EU) regulation. The risk classifications proposed are: «Unacceptable Risk», «High Risk» and «Low Risk», and each classification will have certain particularities. d) Prior authorisation from the IFT will be required to offer AIS in Mexico, even in those cases where AIS is offered free of charge. e) The bill establishes as a penalty for non-compliance up to 10% of the annual income of the infringing entities.” (https://ricardomonrealavila.com/wp-content/uploads/2024/02/Inic_Morena_inteligencia_artificial.pdf; https://basham.com.mx/bill-to-regulate-artificial-intelligence-in-mexico-key-points/) |
|||
| Netherlands | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, the Netherlands needs to implement several key measures, including the following important steps as part of the Netherlands’ national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Peru | Draft bill proposed | International treaty/declaration: 9 August 2024 – ‘Cartagena de Indias Declaration for Governance, the construction of AI ecosystems and the promotion of AI education in an Ethical and Responsible manner in Latin America and the Caribbean’; signed by Peru (alongside 16 other Latin American countries). National legislation: 2 May 2024 – Draft law no. 31814 ‘promoting the use of AI for the economic and social development of the country’ published. |
|
|
International treaty/declaration:
9 August 2024 – ‘Cartagena de Indias Declaration for Governance, the construction of AI ecosystems and the promotion of AI education in an Ethical and Responsible manner in Latin America and the Caribbean’ (the “Declaration”); signed by 17 countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Curazao, Ecuador, Guatemala, Guyana, Honduras, Panama, Paraguay, Peru, Dominican Republic, Suriname and Uruguay. The Declaration is a framework which represents Latin America’s proactive stance in addressing AI challenges while promoting responsible innovation. It emphasises the importance of digital skills development and knowledge sharing across borders. Core Components: • Regional cooperation framework for AI development • Comprehensive governance structures emphasising safety and ethics • Supportive ecosystems for responsible innovation • Educational initiatives to build digital capacity • Strong emphasis on human rights protection • Alignment with international standards and UN initiatives (https://mintic.gov.co/portal/715/articles-383990_recurso_1.pdf).National legislation: 2 May 2024 – Draft law no. 31814 promotes the safe and responsible use of AI for the economic and social development of the country. The draft of 2 May 2024 replaces a shorter draft issued in July 2023 (https://spijweb.minjus.gob.pe/wp-content/uploads/2023/07/L-31814.pdf). The draft regulations provide a high-level framework for the safe and responsible use of AI, including a risk-based approach similar to the EU which classifies AI based systems as “unacceptable”, “high”, “medium” and “low”. The draft also includes provisions supporting governance of AI by the Secretariat of Government and Digital Transformation, charged with directing, evaluating and supervising the use and promotion of the development of AI and emerging technologies (https://cdn.www.gob.pe/uploads/document/file/6273073/5516872-proyecto-reglamento-ley-31814.pdf?v=1714659628). |
|||
| Poland | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Poland needs to implement several key measures, including the following important steps as part of Poland’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Portugal | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Portugal needs to implement several key measures, including the following important steps as part of Portugal’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Romania | Enacted | National legislation NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Romania needs to implement several key measures, including the following important steps as part of Romania’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Russia | Enacted | National legislation: 24 April 2020 – Federal Law No. 123 FZ adopted. 31 July 2020 – Federal Law No. 258 FZ adopted. | |
|
National Legislation:
Russia adopted Federal Law No. 123 FZ on 24 April 2020 which introduced a five year experimental regime for the testing and use of AI in Russia, and in particular Moscow City, and introduced changes to existing laws on the storage and use of personal data. https://cis-legislation.com/document.fwx?rgn=124089 Federal Law No. 258 FZ adopted on 31 July 2020 the “Law on Experimenting with Artificial Intelligence”, introduced the ability for new technologies to be tested in a sandbox environment (https://cis-legislation.com/document.fwx?rgn=126433). |
|||
| Singapore | None | National legislation: None | |
| Slovakia | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Slovakia needs to implement several key measures, including the following important steps as part of Slovakia’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Slovenia | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Slovenia needs to implement several key measures, including the following important steps as part of Slovenia’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| South Korea | Enacted | National legislation: 16 May 2023 – The Bill amending the Intelligent Robot Development and Promotion Act (Bill no. 2116941) adopted (in force on 17 November 2023). 26 September 2024 – Amendment to the Special Act on Punishment of Crimes of Sexual Violence passed. | |
|
National Legislation:
16 May 2023 – the amendment to the Intelligent Robot Development and Promotion Act (Bill no. 2116941) (in force 17 November 2023) establishes operating conditions for outdoor mobile robots in delivery and surveillance sectors. Key provisions include mandatory damage insurance coverage for certified robots and safety certification requirements overseen by the Minister of Trade, Industry, and Energy based on a checklist. The amendment introduces stricter oversight of certification agencies and enhanced anti-fraud measures. (https://elaw.klri.re.kr/eng_mobile/viewer.do?hseq=63030&type=lawname&key=intelligent+robot) 26 September 2024 – The government passed an amendment to the Special Act on Punishment of Crimes of Sexual Violence, imposing prison sentences for the possession or viewing of deepfake sexual content. (https://likms.assembly.go.kr/bill/billDetail.do?billId=PRC_V2K4Y0U9B2E4C2V1M0M5S2Y0W4C3W9 https://likms.assembly.go.kr/bill/billDetail.do?billId=PRC_X2H2N0A8K1I2V0W9N2O2S1Z9F2B0C5 ) |
|||
| Spain | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Spain needs to implement several key measures, including the following important steps as part of Spain’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Sweden | Enacted | National legislation: NA. Compliance with the EU AI Act pending. | |
|
Pending compliance with the EU AI Act
The EU AI Act entered into force on 1 August 2024 and, save where explicitly stated, will become generally applicable from 2 August 2025. To comply with this legislation, Sweden needs to implement several key measures, including the following important steps as part of Sweden’s national AI governance framework: By August 2025: Appoint at least two national authorities – one to oversee compliance and another for market supervision. By August 2026: Create regulatory sandboxes designed to foster AI innovation while ensuring adherence to legal requirements. Any supplemental national legislation required in the future to give effect to certain provisions under the EU AI Act will be added here. |
|||
| Switzerland | None | International treaty/declaration: 5 September 2024 – Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (“AI Convention”) concluded by Council of Europe member states, with the EU, UK, USA and others. National legislation: none | |
|
International treaty/declaration:
On 5 September 2024, the AI Convention, a multilateral international treaty was concluded between EU member states, European Council member states, which includes Switzerland, and other signatories including the US. The treaty aims to ensure that Activities within the lifecycle of AI systems comply with fundamental principles including human rights, democracy and the rule of law (https://rm.coe.int/1680afae3c).
National Legislation: None |
|||
| Thailand | In progress | National legislation: 18 October 2022 – Draft Royal Decree on Business Operations that Use Artificial Intelligence Systems published; 30 March 2023 – Draft Act on the Promotion and Support of AI Innovations published; Draft ETDA Notification regarding AI sandboxes published; Draft ETDA Notification regarding AI Risk Assessments published. | |
|
National Legislation:
18 October 2022 – The Draft Royal Decree on Business Operations that Use Artificial Intelligence Systems sets out a risk-based approach similar to that of the EU’s AI Act, regulating AI systems available to consumers based on their categorisation as “Limited Risk”, “High Risk” and “Unacceptable Risk”. The draft is currently under consideration. 30 March 2023 – The Draft Act on the Promotion and Support of AI Innovations was introduced by the Electronic Transactions Development Agency (ETDA) for the purposes of an AI regulatory framework to support an AI ecosystem. The draft is currently under consideration. In parallel, two further draft regulations were published by the ETDA concerning AI sandboxes (to allow developers to test AI technology in a safe digital environment), and AI risk assessments in connection with AI systems; https://www.law.go.th/listeningDetail?survey_id=MjM4M0RHQV9MQVdfRlJPTlRFTkQ= |
|||
| Turkey | In progress | National legislation: 24 June 2024 – Artificial Intelligence Law (2/2234) introduced. | |
|
National Legislation:
24 June 2024 – The law aims to ensure the safe, ethical and fair use of artificial intelligence technologies, to ensure the protection of personal data, to prevent the violation of privacy rights and to establish a regulatory framework for the development and use of artificial intelligence systems. The law is currently under consideration. https://www.tbmm.gov.tr/Yasama/KanunTeklifi/e21539a0-888a-4500-81be-01904a918c53 |
|||
| UK | None | International treaty/declaration: 5 September 2024 – UK signed the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (“AI Convention”) National legislation: There is currently no national legislation regulating AI in the UK. |
|
|
International treaty/declaration: 5 September 2024 – The UK became a signatory to the AI Convention, a multilateral international treaty (other signatories include the EU and the US) which aims to ensure that Activities within the lifecycle of AI systems comply with fundamental principles including human rights, democracy and the rule of law (https://rm.coe.int/1680afae3c).
National Legislation:
National legislation: There is currently no national legislation regulating AI in the UK, though on 17 July plans were announced introduce legislation that would focus on regulating developers(https://www.gov.uk/government/speeches/the-kings-speech-2024). A draft bill is awaited. |
|||
| United States | State legislation only | International treaty/declaration: 5 September 2024 – Council of Europe Framework Convention on AI and Human Rights. National legislation: For Federal and State legislation, click “Show analysis”. |
|
|
International Treaty/Declaration:
The AI Convention aims to ensure that activities within the lifecycle of AI systems comply with fundamental principles including human rights, democracy and the rule of law (https://rm.coe.int/1680afae3c).
Federal Legislation:
Currently, there is no comprehensive federal law in the United States specifically regulating AI as a whole. A significant number of draft bills have been introduced before the 118th Congress, such as H.R.7913 Generative AI Copyright Disclosure Act of 2024, which would require developers of AI technologies to identify the materials used to train their systems. However, on 20 January 2025 President Donald Trump rescinded the Biden Administration’s Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. This executive order, issued in October 2023, had adopted a broad definition of AI, encompassing any “machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.” The order extended beyond just generative AI to include a wide range of AI systems and aimed to establish safeguards for their development and use. Its revocation suggests that the US may steer clear of federal level regulation going forward, leaving only state-level legislation and industry-led standards.
State Legislation:
Various states have passed bills to regulate AI across a multitude of areas such as individual privacy, deepfake technology, AI data transparency, protection of minors. Some of these are outlined below as examples. (We are working on a more comprehensive tracker to follow US State legislation more closely.)
California
24 May 2024 – “Crimes: Distribution of intimate images” California Senate Bill SB 926 passed. This legislation expands existing laws around the distribution of intimate images to include artificially created content. Key provisions include: • Makes it a crime for a person 18 or older to intentionally create and distribute photorealistic or computer-generated intimate images of an identifiable person without their consent • Applies to images depicting intimate body parts or sexual acts that a reasonable person would believe to be authentic • The distribution must be done knowing it will likely cause serious emotional distress to the depicted person Florida
6 June 2023 – Florida Digital Bill of Rights Senate Bill SB 262 passed. The law became effective from 1 July 2023, and provides data protections for Florida residents. Notably, it allows consumers to opt out of AI-based data collection and use by “controllers” as defined within the text. 29 February 2024 – Political Deepfake Regulation Florida House Bill HB 919 passed. Effective from 1 July 2024, the legislation aims to address AI use in political advertisements in an attempt to combat use of “deepfakes” in political advertising. |
|||
| Vietnam | In progress | National legislation: 2 July 2024 – Draft Digital Technology Industry Law published. | |
|
National Legislation: 2 July 2024 – The draft Digital Technology Industry Law aims to establish a legal framework to promote the development of multiple digital technologies in Vietnam, including AI (Chapter 6), and providing a mechanism for controlled experiments of AI (Chapter 4) (https://mic.gov.vn/luat-cong-nghiep-cong-nghe-so-se-tao-moi-truong-phap-ly-thuan-loi-cho-doanh-nghiep-cong-nghe-so-but-pha-197240703111123061.htm ) | |||