The agreement also includes obligations for AI systems that are classified as “high risk” and a “two-tier” system of guardrails for “general” AI systems. Penalties for non-compliance can lead to fines ranging from €35 million or 7% of global turnover to €7.5 million or 1.5 % of turnover, depending on the infringement and size of the company. The law will be phased in, with full implementation expected by 2026. The EU believes the law will provide legal and technical certainty for developers, startups, and future scale-ups.
Key takeaways:
- European Union lawmakers have reached a political agreement on a risk-based framework for regulating artificial intelligence (AI), which includes a total prohibition on the use of AI for certain activities such as biometric categorisation systems and untargeted scraping of facial images.
- The agreement also includes obligations for AI systems that are classified as “high risk” and a “two-tier” system of guardrails to be applied to “general” AI systems.
- Penalties for non-compliance can lead to fines ranging from €35 million or 7% of global turnover to €7.5 million or 1.5 % of turnover, depending on the infringement and size of the company.
- The full force of the EU’s AI Act may not be felt until 2026, with a phased entry into force after the law is adopted.