EU AI Act — Political agreement¶
★★★★★ Basic
Development line: project:eu-ai-act · thread political-agreement
Last event: 2023-12-10 · 1 dated since 2023-12-10 · Researched: 2026-09-05 · confidence: high
What it is¶
EU AI Act is Regulation (EU) 2024/1689 for companies that build, sell, integrate or use AI in the EU. It covers prohibited practices, high-risk systems, transparency and general-purpose AI models. Annex III high-risk rules apply from 2 December 2027; product-embedded Annex I rules apply from 2 August 2028. We classify the system and our role before treating a policy or model card as compliance.
Development line¶
- 2023-12-10 — EU institutions reached a political agreement on the AI Act. On 2023-12-10, EU institutions reached a political agreement on the EU AI Act. This moved negotiation toward a formal regulatory framework. The available links do not establish the final legal text, adoption sequence, or entry-into-force date. Public discussion and redistribution followed across the additional links.
What changed¶
- 2023-12-10 — A provisional political agreement set the negotiated direction.
- 2024-06-13 — Regulation (EU) 2024/1689 was adopted.
- 2024-08-01 — The Act entered into force.
- 2025-02-02 — Prohibited-practice and AI-literacy provisions began to apply.
- 2025-08-02 — Governance and GPAI-model obligations began to apply.
- 2026-07-27 — The AI Omnibus changed the high-risk timetable.
- 2026-08-02 — Most provisions and Article 50 transparency obligations became applicable.
- 2027-12-02 — Annex III high-risk rules are scheduled to apply.
- 2028-08-02 — Annex I product-embedded high-risk rules are scheduled to apply.
How to use this¶
From 2023-12-10, map affected AI systems and governance processes against the agreed EU AI Act direction. Treat final obligations and timelines as unconfirmed until the text is enacted.
- Identify your role and EU exposure; use the official checker as a starting point, not legal advice. — https://ai-act-service-desk.ec.europa.eu/en/eu-ai-act-compliance-checker
- Classify the system against prohibited practices, transparency duties and high-risk categories, retaining evidence and the applicable date. — https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai
- If you provide a GPAI model, maintain authority and downstream documentation, copyright policy and training-content summary. — https://digital-strategy.ec.europa.eu/en/faqs/guidelines-obligations-general-purpose-ai-providers
- For interactive or generative systems, implement applicable Article 50 notices and machine-readable output marking before market placement or service use. — https://digital-strategy.ec.europa.eu/en/faqs/transparency-obligations-under-article-50-ai-act
Best practices¶
- Maintain a versioned inventory of systems, purpose, suppliers, EU exposure, role, classification evidence and review date. — https://digital-strategy.ec.europa.eu/en/faqs/guidelines-obligations-general-purpose-ai-providers
- Use Commission guidance and the checker as operational aids; obtain tailored legal advice where the classification has material consequences. — https://ai-act-service-desk.ec.europa.eu/en/eu-ai-act-compliance-checker
- Use an approved code of practice or retain evidence that an alternative means is adequate. — https://digital-strategy.ec.europa.eu/en/faqs/transparency-obligations-under-article-50-ai-act
Superseded by this¶
- 2023-12-10 political agreement — superseded as the operative source by Regulation (EU) 2024/1689, adopted 2024-06-13 and published 2024-07-12.
- Pre-2026 high-risk timetable — superseded 2026-07-27: Annex III rules now apply from 2027-12-02 and Annex I rules from 2028-08-02.
Still unknown¶
- We did not rely on the supplied source and social-media URLs because their content was unavailable for verification.
Sources¶
| source | title | read |
|---|---|---|
| https://www.consilium.europa.eu/en/press/press-releases/2023/12/09/artificial-intelligence-act-council-and-parliament-strike-a-deal-on-the-first-worldwide-rules-for-ai/ | Artificial intelligence act: Council and Parliament strike a deal on the first rules for AI in the world | 2026-09-05 |
| https://techcrunch.com/2023/12/08/eu-ai-act-political-deal/ | EU lawmakers bag late night deal on 'global first' AI rules | 2026-09-05 |
| https://eur-lex.europa.eu/eli/reg/2024/1689/2024-07-12/eng | Regulation (EU) 2024/1689 — Artificial Intelligence Act | 2026-09-05 |
| https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai | AI Act | 2026-09-05 |
| https://ai-act-service-desk.ec.europa.eu/en/eu-ai-act-compliance-checker | EU AI Act Compliance Checker | 2026-09-05 |
| https://digital-strategy.ec.europa.eu/en/faqs/guidelines-obligations-general-purpose-ai-providers | Guidelines on obligations for General-Purpose AI providers | 2026-09-05 |
| https://digital-strategy.ec.europa.eu/en/faqs/transparency-obligations-under-article-50-ai-act | Transparency obligations under Article 50 of the AI Act | 2026-09-05 |
| https://digital-strategy.ec.europa.eu/en/news/ai-omnibus-enters-force | AI Omnibus enters into force | 2026-09-05 |
Agent brief¶
- Subject:
project:eu-ai-act, threadpolitical-agreement, 1 dated events 2023-12-10 → 2023-12-10. - Practical note: From 2023-12-10, map affected AI systems and governance processes against the agreed EU AI Act direction. Treat final obligations and timelines as unconfirmed until the text is enacted.
- Confidence: high. Dated supersedes above are the authority for what is obsolete.