On this page
Educational content, not legal advice. This explains publicly available material as of the publication date. Regulatory requirements change, so confirm how they apply to your organisation with qualified counsel.
Regulation (EU) 2024/1689, the AI Act, is the first comprehensive horizontal law on artificial intelligence from a major jurisdiction. It entered into force on 1 August 2024 and applies in stages. That timetable changed in 2026, when Regulation (EU) 2026/1744, the Digital Omnibus on AI, entered into force on 27 July 2026 and postponed the main high-risk deadlines. Guidance written before then may quote dates that no longer apply.
This is an educational overview for business and technology leaders, based on the European Commission's published materials as of the date above. It is not legal advice, and your legal counsel should confirm how the Act applies to you.
The current timeline
| Date | What applies |
|---|---|
| 1 August 2024 | The Act enters into force |
| 2 February 2025 | Prohibited practices, definitions and AI literacy provisions |
| 2 August 2025 | Rules for general-purpose AI models, and governance arrangements |
| 2 August 2026 | Most remaining rules, including Article 50 transparency duties, and the start of enforcement for the rules that apply |
| 2 December 2026 | New prohibition on systems generating non-consensual sexual deepfakes and child sexual abuse material. Transition ends for certain existing systems under Article 50(2) |
| 2 December 2027 | Rules for high-risk systems in Annex III areas, such as employment and education |
| 2 August 2028 | Rules for high-risk AI embedded in regulated products under Annex I |
The Commission's AI Act Service Desk timeline is the best place to check for changes. The postponement is not a pause: prohibitions, general-purpose AI rules and transparency duties are already in application.
The risk-based approach
The Commission describes four levels of risk.
- Unacceptable risk. These practices are banned, such as social scoring.
- High risk. These systems can seriously affect health, safety or fundamental rights, and they carry strict obligations such as risk management, documentation and human oversight.
- Transparency risk. These carry disclosure duties, such as telling people they are talking to a machine and labelling deepfakes.
- Minimal or no risk. The Act sets no rules here, and the Commission says most AI systems in use in the EU fall into this level.
General-purpose AI models have their own set of obligations, which fall mainly on the model's provider.
Your role determines your duties
Enterprises often assume the Act is mainly a problem for AI vendors. In practice, many large organisations will hold more than one role across their portfolio.
Providers
A provider develops an AI system or has it developed, and places it on the market or puts it into service under its own name. Providers of high-risk systems carry the heaviest duties, including a risk management system, data governance, technical documentation, logging, a quality management system and conformity assessment.
Deployers
A deployer uses an AI system under its own authority in a professional context. Deployers of high-risk systems have duties too, including using the system according to its instructions, assigning competent human oversight, monitoring operation and keeping logs. Some must also carry out a fundamental rights impact assessment.
When a deployer becomes a provider
Article 25 can shift provider obligations to an organisation that puts its own name on a high-risk system, substantially modifies one or changes a system's intended purpose so that it becomes high-risk. Teams that customise vendor AI should check this carefully.
Reach beyond the EU
The Act can apply to organisations established outside the EU, including where a system's output is used in the Union. Enterprises in India, the Middle East and North America should not assume that being headquartered elsewhere places them out of scope.
What to do now
- Build an inventory with an EU lens. Record each AI system, including vendor features, and whether it, or its output, touches the EU.
- Confirm nothing is prohibited. These rules have applied since February 2025. Review them against the Article 5 text, including the amendments made in 2026.
- Classify your systems and your role. Check uses against Annex III, and record your reasoning. Article 6(3) allows some Annex III systems to fall outside the high-risk category under defined conditions, so check the current text.
- Address transparency. Article 50 duties now apply, so review chatbots, synthetic content and deepfakes.
- Keep AI literacy on the agenda. The Omnibus amended Article 4. Published commentary describes the duty as now being to take measures supporting AI literacy, so ask your counsel to confirm what this means for you.
- Use the extra time for high-risk systems. Risk management, data governance and technical documentation take time to embed. Starting them early also avoids building them under deadline pressure.
- Review supplier contracts. As a deployer, you will depend on vendors for instructions, documentation and cooperation.
- Connect to existing compliance work. The Act does not replace the GDPR. Align AI reviews with data protection impact assessments and your broader security programme.
Enforcement and penalties
According to the Commission's enforcement page, the highest tier of penalties, for prohibited practices, reaches up to €35 million or 7% of total worldwide annual turnover. Breaches of general-purpose AI obligations fall under a lower tier of up to €15 million or 3%. National authorities and the Commission's AI Office share enforcement, depending on the matter.
What to watch next
Three developments are worth tracking: harmonised technical standards for high-risk systems, further Commission guidance, and how national authorities approach enforcement in practice. Assign someone to monitor them, and treat your compliance assumptions as something to review regularly and not as a one-time decision.
Sources
- Regulation (EU) 2024/1689 (AI Act), EUR-Lex
- Regulation (EU) 2026/1744 (Digital Omnibus on AI), EUR-Lex
- AI Act Service Desk: Timeline for the implementation of the EU AI Act, European Commission
- AI Act policy page, European Commission
- The enforcement framework of the AI Act, European Commission
- Regulation (EU) 2016/679 (GDPR), EUR-Lex