AI Act (Tiered Penalties)
EU AI Act fine structure: up to €35M or 7% of global turnover for prohibited practices; up to €15M or 3% for other obligation breaches.
What AI Act (Tiered Penalties) means
The EU AI Act (Regulation 2024/1689) introduces a tiered penalty structure for non-compliance, modelled loosely on the GDPR fine framework. The highest tier — up to €35 million or 7% of total global annual turnover (whichever is higher) — applies to violations of the Act's prohibited AI practices (e.g., social scoring, manipulative cognitive techniques, real-time remote biometric identification in public spaces). A middle tier of up to €15 million or 3% applies to violations of other AI Act obligations, including transparency requirements for generative AI systems (Article 50) and high-risk AI system requirements. A lower tier of up to €7.5 million or 1.5% applies to supplying incorrect information to authorities. Phased application begins August 2024 with prohibited practices, with full enforcement of generative-AI transparency obligations from 2 August 2026.