The European Commission states that the AI Act entered into force in August 2024 and became generally applicable on 2 August 2026, with exceptions and extended transition dates for specified high-risk systems.

The consequential change is cultural: organisations must increasingly demonstrate how AI is governed, not merely declare that it is responsible.

Why this matters

The Act turns broad principles into operating obligations. For leaders, the challenge moves toward inventories, risk classification, technical documentation, transparency, human oversight and evidence that controls work in practice.

What AIAF is watching