Glossary
Updated
What is the EU AI Act?
The EU AI Act, Regulation (EU) 2024/1689, is the European Union's law on artificial intelligence. It entered into force on August 1, 2024, and applies in stages. It regulates AI by what it's used for: the higher the risk to health, safety or fundamental rights, the stricter the rules. In 2026 the Digital Omnibus, Regulation (EU) 2026/1744, amended it. It moved the high-risk dates, softened the AI literacy duty and added two bans.
Who does the EU AI Act apply to?
Mainly providers, who develop an AI system or general-purpose AI model, or have one built, and offer or use it under their own name, and deployers, who use an AI system under their authority at work. It covers providers placing AI on the EU market wherever they're based, deployers established in the EU, and providers and deployers outside the EU whose AI output is used in the EU. Importers and distributors have duties too. A deployer can become a provider under Article 25, for example by putting its name on a high-risk system or changing a system's purpose so it becomes high-risk.
What are the risk tiers?
Prohibited practices under Article 5 are banned, such as social scoring and emotion recognition at work. High-risk systems are either listed in Annex III, such as CV screening, credit scoring and life and health insurance pricing, or built into products covered by EU safety law under Annex I. They carry the heaviest duties: risk management, data governance, documentation, logging, human oversight and conformity assessment for providers, and use as instructed, trained oversight and log keeping for deployers.
Systems with transparency duties under Article 50, such as chatbots and deepfakes, must tell people they're dealing with AI, or that content is AI-made. Everything else is minimal risk, with no new duties beyond AI literacy. General-purpose AI models have their own set of rules.
When do the rules apply?
The bans and the AI literacy duty have applied since February 2, 2025, and the duties for general-purpose AI models since August 2, 2025. Transparency duties under Article 50 have applied since August 2, 2026. Two new bans, on AI that makes nonconsensual intimate deepfakes or child sexual abuse material, apply from December 2, 2026. High-risk duties apply from December 2, 2027, for Annex III uses and from August 2, 2028, for AI in regulated products. Older high-risk AI used by public authorities has until August 2, 2030.
What are the fines under the EU AI Act?
Fines reach €35 million or 7% of worldwide annual turnover, whichever is higher, for a banned practice. Most other breaches, including provider, deployer and transparency duties, carry up to €15 million or 3%. Giving authorities incorrect or misleading information carries up to €7.5 million or 1%. Smaller companies face the lower of the two amounts. Since the Omnibus, small mid-caps do too, except for banned practices.
Our EU AI Act compliance guide sets out each duty, who it falls on and how Enzai helps. To work out which tier a system falls into, see the EU AI Act risk classification guide.
Example
A US company sells an AI tool that screens job applications to employers in Germany. Recruitment is listed in Annex III, so the tool is high-risk, and the company is its provider even though it has no EU office. By December 2, 2027, it needs a risk management system, technical documentation, a conformity assessment, CE marking, registration in the EU database and an authorized representative in the EU. Its German customers are deployers. They must use the tool as instructed, give oversight to trained staff, tell applicants it is used in decisions about them, and keep its logs for at least six months.
