EU Regulation
Artificial Intelligence Act (AI Act)
Sets risk-based requirements for AI providers and deployers in the EU, from banned practices to high-risk systems and general-purpose models.
- Feb 2025 prohibited practices banned and AI literacy measures due
- Aug 2025 obligations for general-purpose AI models apply
- €35m or 7% of worldwide turnover: the maximum fine for prohibited practices
What Is AI Act?
The EU AI ActArtificial Intelligence Act was enacted to mitigate risks associated with artificial intelligenceMachine-based systems that infer from the input they receive how to generate outputs such as predictions, content, recommendations or decisions. The EU AI Act defines an AI system in these terms, covering systems that operate with varying levels of autonomy, and regulates them according to risk. by establishing robust standards for safety, transparency, and accountability. It aims to prevent discriminatory outcomes, protect fundamental rights, and promote trust in AIArtificial Intelligence technologies, while simultaneously fostering innovation and competitive growth across the European digital economy.
The EU Artificial Intelligence ActAn EU regulation governing the supply and use of artificial intelligence systems and general-purpose models. General-purpose models can perform a wide range of tasks and be integrated into other systems. The act restricts certain uses and sets requirements for specified high-risk uses, transparency and general-purpose models. High-risk uses are categories identified by the act because of their potential effects on health, safety or fundamental rights. is the EU regulation on artificial intelligenceMachine-based systems that infer from the input they receive how to generate outputs such as predictions, content, recommendations or decisions. The EU AI Act defines an AI system in these terms, covering systems that operate with varying levels of autonomy, and regulates them according to risk.. It sorts AIArtificial Intelligence systems by risk, bans a small set of practices, and sets requirements for high-risk systems and general-purpose AIAn AI model with significant generality that can competently perform a wide range of distinct tasks, as defined in AI Act Article 3(63). Article 3(66) separately defines a general-purpose AI system based on such a model that can serve a variety of purposes. Provider requirements depend on the model and applicable transitional provisions, including Article 111(3) for models already on the market. models, with obligations phasing in from February 2025.
The regulation aims to ensure transparency, accountability, and fairness, with fines of up to EUR 35 million or 7% of worldwide annual turnover for prohibited practices (Article 99(3)). It is designed to foster innovation while protecting fundamental rights and ethical standards across industries.
- Type
- Regulation
- Issued by
- European Parliament and the Council of the European Union
- Adopted
- 13 June 2024
- In force
- 1 August 2024
- Applies from
- 2 August 2026
- Status
- Applies in stages
- Countries
- All 27 EU member states
Verified by
Ahmet Ucar
Principal Consultant (CISSPCertified Information Systems Security Professional, CISMCertified Information Security Manager, CISACertified Information Systems Auditor, CIPP/ECertified Information Privacy Professional/Europe)
Last updated 18 November 2026
Does It Apply to You?
Answer once on your sectors, size and countries. You get the result per country, with the articles behind it.
What the AI Act Requires
Article 5(1)
Prohibited AI practices
Do not place on the market, put into service or use AI that manipulates people or exploits their vulnerabilities causing, or likely to cause, significant harm, scores people socially leading to unjustified or out-of-context detriment, predicts crime from profiling alone, builds facial recognition databases by untargeted scraping, infers emotions at work or in education other than for medical or safety reasons, or categorises people biometrically to infer sensitive traits. Real-time remote biometric identification in public spaces for law enforcement is banned outside narrow exceptions.
Article 6(1), (1a) to (1c), (2)
Classify high-risk AI systems
Treat an AI system as high-risk where it is a safety component of a product, or is itself a product, under Annex I legislation that requires third-party conformity assessment, or where it falls in an Annex III use case. A third-party assessment required only for risks other than health and safety does not count. AI used only for non-safety assistance, optimisation, efficiency, convenience or quality control is not a safety component, unless its failure would endanger health and safety.
Article 9(1), (2)
Risk management system
Establish, implement, document and maintain a risk management system for each high-risk AI system. Run it as a continuous, iterative process across the whole lifecycle, with regular systematic review and updating.
Article 26(5)
Monitor use and report problems
Monitor operation on the basis of the instructions. If use may present a risk, inform the provider or distributor and the market surveillance authority without undue delay and suspend use. If you identify a serious incident, immediately inform the provider first, then the importer or distributor and the market surveillance authorities.
Show All 6 Requirements
Article 50(1), (5)
Disclose interaction with AI
Design systems that interact directly with people so those people are told they are dealing with AI, unless this is obvious to a reasonably well-informed person. Give the information clearly at the latest at the first interaction, in an accessible form.
Article 99(3), (6)
Fines for prohibited practices
Breach of the Article 5 prohibitions carries fines of up to EUR 35 million or 7% of total worldwide annual turnover, whichever is higher. For SMEs and start-ups the lower of the two applies.
More Requirements in Related Documents
- KI-lovenRule · 4 requirements
- Decreet 16 oktober 2026Rule · 3 requirements
- No. 31/2026Rule · 5 requirements
| Covered entities | Kind | Article | Where defined |
|---|---|---|---|
| ProviderProvider Under the AI Act | Role | Article 3(3); Article 2(1)(a) | Develops an AIArtificial Intelligence system or general-purpose AIAn AI model with significant generality that can competently perform a wide range of distinct tasks, as defined in AI Act Article 3(63). Article 3(66) separately defines a general-purpose AI system based on such a model that can serve a variety of purposes. Provider requirements depend on the model and applicable transitional provisions, including Article 111(3) for models already on the market. model, or has it developed, and places it on the market or puts it into service under its own name or trademark. |
| ProviderProvider Under the AI Act of a high-risk AI systemAn AI system classified as high-risk under AI Act Article 6, either through the Annex I product-safety route or an Annex III use case. Certain Annex III systems may fall outside that classification under Article 6(3) if its conditions are met, while Annex III systems performing profiling of natural persons remain high-risk. Classification determines the applicable conformity and compliance requirements. | Role | Article 6; Article 16 | AIArtificial Intelligence systems classified as high-risk under Article 6 and Annex III; carries the obligations in Article 16. |
| DeployerA natural or legal person, public authority, agency or other body that uses an AI system under its authority, other than for personal non-professional activity. Deployers of high-risk AI systems must follow the instructions for use, ensure human oversight and monitor the system. | Role | Article 3(4); Article 26 | Uses an AIArtificial Intelligence system under its authority, except for personal non-professional activity. |
| Importer | Role | Article 3(6); Article 23 | Established in the Union and places on the market an AIArtificial Intelligence system bearing the name or trademark of a person established in a third country. |
| Distributor | Role | Article 3(7); Article 24 | Makes an AIArtificial Intelligence system available on the Union market, other than the providerProvider Under the AI Act or importer. |
| Authorised representative | Role | Article 3(5); Article 22 | Appointed in the Union by a providerProvider Under the AI Act established in a third country. |
| Product manufacturers | Role | Article 2(1)(e) | Place on the market or put into service an AIArtificial Intelligence system together with their product and under their own name or trademark. |
| ProviderProvider Under the AI Act of a general-purpose AIAn AI model with significant generality that can competently perform a wide range of distinct tasks, as defined in AI Act Article 3(63). Article 3(66) separately defines a general-purpose AI system based on such a model that can serve a variety of purposes. Provider requirements depend on the model and applicable transitional provisions, including Article 111(3) for models already on the market. model | Role | Article 3(3) and (63); Article 51 | Additional obligations where the model is classified as having systemic riskUnder the AI Act, a risk specific to the high-impact capabilities of general-purpose AI models that can have a significant impact on the EU market, or foreseeable negative effects on public health, safety, security, fundamental rights or society, that can spread at scale. Providers of such models face extra duties such as evaluations and incident reporting. (Articles 51 and 55). |
The EU AI ActArtificial Intelligence Act applies to all providersProvider Under the AI Act, developers, importers, and users of artificial intelligenceMachine-based systems that infer from the input they receive how to generate outputs such as predictions, content, recommendations or decisions. The EU AI Act defines an AI system in these terms, covering systems that operate with varying levels of autonomy, and regulates them according to risk. systems operating within the European Union. It covers entities deploying high-risk AIArtificial Intelligence applications in sectors such as healthcare, transportation, finance, and public administration, ensuring that both large corporations and small enterprises adhere to stringent safety and transparency requirements.
2 days
Serious incident report Article 73
When: Within 2 days of becoming aware for a widespread infringement or critical infrastructure disruption; 10 days for a death; 15 days otherwise
Every year
Annual report on post-remote biometric identification Article 26(10)
Every 5 years
Renew certificate for Annex I systems Article 44(2)
Every 4 years
Renew certificate for Annex III systems Article 44(2)
| Applies to | Maximum fine |
|---|---|
| Article 99(3) prohibited practices | EUR 35 million or 7% of worldwide (group) annual turnover, whichever is higher |
| Article 99(4) operator and notified bodyA conformity assessment body that a Member State has notified to the Commission as competent to carry out third-party assessment under EU legislation. Whether a manufacturer or provider must involve one depends on the applicable act, system or product classification and conformity assessment route; not every high-risk AI system requires a notified body. obligations | EUR 15 million or 3% of worldwide (group) annual turnover, whichever is higher |
| Article 99(5) incorrect, incomplete or misleading information | EUR 7.5 million or 1% of worldwide (group) annual turnover, whichever is higher |
| Article 101 general-purpose AIAn AI model with significant generality that can competently perform a wide range of distinct tasks, as defined in AI Act Article 3(63). Article 3(66) separately defines a general-purpose AI system based on such a model that can serve a variety of purposes. Provider requirements depend on the model and applicable transitional provisions, including Article 111(3) for models already on the market. model providersProvider Under the AI Act (fined by the Commission) | EUR 15 million or 3% of worldwide (group) annual turnover, whichever is higher |
Non-compliance with the EU AI Act may result in severe penalties, including administrative fines of up to EUR 35 million or 7% of worldwide annual turnover, whichever is higher, for prohibited practices (Article 99(3)). In addition, businesses may face operational restrictions, reputational damage, and potential bans on deploying high-risk AI systems. Regulatory authorities may also impose corrective measures to ensure compliance, emphasizing the importance of adherence to the established standards to protect user rights and maintain ethical practices.
Latest AI Act Updates
All AI Act Updates-
19 November 2026DesignationBelgium
Flanders Makes Its Media Regulator the AI Act Authority for Deepfakes in Broadcasting
A Flemish decree published on 19 November 2026 makes the Vlaamse Regulator voor de Media the AI Act authority for deepfake disclosure by broadcasters, with fines of up to EUR 15 million or 3% of turnover.
-
16 November 2026Bill introducedEstonia
Estonia Introduces a Bill Creating an Agent Code for AI Agents Acting for People
Estonia's Government introduced 1092 SE on 16 November 2026. It would create an agent code that links actions taken by AI agents to a person, and allow such actions in the state information system only with that code.
-
16 November 2026ConsultationEuropean Union
Commission Seeks Feedback on AI Act Fundamental Rights Impact Assessment Template
The European Commission has published a draft template for the fundamental rights impact assessment under Article 27 of the AI Act. Feedback is open until 15 January 2027. Deployers who must carry out the assessment have to comply from 2 December 2027.
AI Act Milestones Since 2024
- AI sandbox implementing regulation published in the Official JournalThe Official Journal published Commission Implementing Regulation (EU) 2026/2487.
- Commission adopts AI sandbox implementing regulationThe adoption date is for Commission Implementing Regulation (EU) 2026/2487.
- AI Act starts to apply in Northern Ireland (Articles 4, 5 and 50(1))Applies to physical products with closely integrated AIArtificial Intelligence placed on the Northern Ireland market, from the Windsor Framework Joint Committee decision.
- Implementing Regulation (EU) 2026/1755 on GPAI proceedings in forceTwentieth day after publication in OJ L 2026/1755 of 21 July 2026.
- Transparency duties apply, the general date of applicationPeople must be told when they deal with AIArtificial Intelligence, and deepfakesDeepfake must be labelled. The Digital Omnibus on AI (Regulation (EU) 2026/1744) moved the high-risk requirements to 2 December 2027 and 2 August 2028.
- AI Omnibus enters into forceRegulation (EU) 2026/1744 postpones the high-risk requirements and adds new bans on AIArtificial Intelligence-generated intimate and abuse imagery.
- General-purpose AI requirements, governance and penalties applyProviders of general-purpose AIAn AI model with significant generality that can competently perform a wide range of distinct tasks, as defined in AI Act Article 3(63). Article 3(66) separately defines a general-purpose AI system based on such a model that can serve a variety of purposes. Provider requirements depend on the model and applicable transitional provisions, including Article 111(3) for models already on the market. models must meet their duties, and member states must have named their authorities.
- Prohibited practices and AI literacy applyProhibited AI practices must have stopped, and providers and deployers must take measures to support the AI literacy of staff who work with AI (Article 4).
- Entered into forceThe requirements apply in stages from this date.
- Adopted by the European Parliament and the CouncilRegulation (EU) 2024/1689 sets requirements by risk, for those who provide and those who use AIArtificial Intelligence systems.
How We Help With the AI Act
Start with an assessment, or have us run the work.
Learn More
Contact us today to learn how our compliance solutions can help you navigate the EU AI ActArtificial Intelligence Act, ensuring your AIArtificial Intelligence systems meet all regulatory standards while driving innovation.