EU AI Act: The Compliance Timeline Every AI Builder Should Know

By Darus S Alfatah · 2026-10-24 · Regulasi · 11 menit

EU AI Act: The Compliance Timeline Every AI Builder Should Know

Author note: Even outside Europe, these rules affect products reaching European users. We start with risk mapping.

Focus of this article

This article focuses on the timeline and obligations of the EU AI Act. For Indonesian AI regulation and ethics, see the companion article.

Read also: AI Regulation in Indonesia: PDP Law, Ethics, and What Startups Need

The First Comprehensive AI Law

The European Union has enacted the AI Act (Regulation (EU) 2024/1689), the first broad legal framework that regulates AI by risk level. It applies in stages, and it covers anyone placing AI systems on the European market — including companies outside Europe that serve EU users.

For Indonesian developers the question is simple: can EU users access your product? If yes, some obligations may apply, especially if you provide general-purpose models or systems used in high-risk domains.

Four Risk Tiers

  • Unacceptable — prohibited, such as systems that manipulate behaviour harmfully or public social scoring.
  • High risk — requires risk management, technical documentation, logging, transparency, and human oversight. Examples: recruitment, credit scoring, product safety components.
  • Limited risk — transparency duties, such as telling users they interact with AI or labelling synthetic content.
  • Minimal risk — most everyday applications; no extra obligations beyond voluntary practice.

Rollout Timeline

The regulation entered into force on 1 August 2024. Prohibitions and AI literacy duties applied from 2 February 2025. Obligations for general-purpose AI (GPAI) models and governance applied from 2 August 2025. High-risk system obligations apply by 2 August 2026 at the latest, while products already regulated under older rules have a longer transition until 2027.

For many teams the most relevant deadlines now are transparency duties and GPAI compliance for those who provide models, not just use them.

Practical Obligations for API Consumers

If you only use third-party models for internal or consumer apps, the burden is lighter but not zero:

  • Chatbot transparency — tell users they are talking to AI.
  • Synthetic content labelling — deepfakes and synthetic media need clear labels.
  • Documentation — keep records of models, versions, and intended use for audits.
  • Risk classification — decide whether your use case is high risk (for example automated CV screening).

Comparing With Indonesia

Indonesia has no dedicated AI law yet. What applies today is Law No. 27 of 2022 on Personal Data Protection, the Ministry of Communication’s AI ethics circular, and a national AI policy that keeps evolving. The principles are similar — transparency, accountability, data protection — but enforcement and penalties are lighter than under the AI Act.

For products serving global users, the safest approach is to adopt the higher standard so you do not need two separate compliance programmes.

Start Now

  • Build an AI system register: every AI feature, model used, and data processed.
  • Classify risk per feature, not per company.
  • Prepare internal model cards and brief technical documentation.
  • Apply human oversight to decisions that affect people.
  • Provide a contact and process for user complaints about AI decisions.

Supply Chain Effects

Documentation obligations flow downstream: model vendors must give customers enough information to meet their own duties. In procurement, security questionnaires and model documents become part of the contract rather than optional attachments. Vendors without documentation will struggle to sell to EU clients even when their own product is not high risk.

Conclusion

Compliance is cheaper when treated as product documentation rather than legal overhead. Teams that record what they build, why, and with which model will pass audits with far less friction.

Sources & References

An example of risk mapping: Record where AI is used, who is affected, and what data is processed. This simple document eases compliance later.

Frequently Asked Questions

Is this regulatory discussion legal advice?

No. It is informational for developers; consult a legal expert for official compliance.

How does AI regulation affect developers?

It generally concerns transparency, data protection, and responsible use — good principles to adopt early.

How do I apply these principles in practice?

Start with safe practices: do not send sensitive data to models without permission, and understand the provider’s privacy policy.

Sources & References

About the author

Darus S Alfatah — Founder & Writer

Darus S Alfatah is the founder of BankAI. He builds and operates the LLM API gateway that serves Indonesian developers, and writes about LLM integration, gateway architecture, and using AI practically and cost-effectively. Every guide is tested in a real environment — including code, configuration, and token calculations — before it is published, and updated when provider documentation or pricing changes.

Editorial policy

Related articles

🚀 Ready to Use the All-in-One API Key?

Sign up, top up with QRIS, and start building with ChatGPT, Claude, DeepSeek and every other model through one key.

🔑 SIGN UP NOW · Pricing