As artificial intelligence continues to grow rapidly, the European Union has introduced the EU AI Act, the world’s first comprehensive regulation on AI.

The regulation officially entered into force in 2024 and will be implemented in phases over the next few years. Businesses that develop or use AI-related products are encouraged to start paying attention to the timeline and potential impact.

📅 Key Timeline (Simplified)

2024

• August 1, 2024 – Regulation enters into force
• Transition period begins (approx. 2 years)

👉 At this stage:
Preparation phase. Most obligations are not yet enforced.

---

2025 (First wave of enforcement)

• February 2, 2025 – Prohibited AI practices banned
  • Examples: social scoring, manipulative AI systems

👉 In short:
The most harmful AI uses are banned first

---

• August 2, 2025 – Generative AI requirements begin
  • Transparency obligations
  • Disclosure of training data (where applicable)
  • Copyright compliance

👉 Focus:
General-purpose AI (e.g. ChatGPT-like models) comes under regulation

---

⭐⭐2026 (Main obligations apply)

• August 2, 2026 – Most AI rules become fully applicable
  • High-risk AI requirements (risk management, documentation, testing, etc.)

👉 This is the key milestone for most businesses

---

2027 (Extended deadlines for certain sectors)

• Additional time for AI embedded in regulated products
  (e.g. medical devices, machinery)

👉 Industry-specific AI may have longer timelines

---

🧭 What Should Businesses Do?

You should start paying attention if your product or service:

• Uses AI (e.g. recommendation systems, chatbots, image recognition)
• Integrates generative AI tools
• Impacts decision-making, safety, or user rights
  (e.g. finance, hiring, healthcare)

For further inquiries, please contact:
Email：Charles.liao@theonelab.co
Phone：(02)8601-2828