European Union member states have begun the next phase of enforcing the AI Act’s obligations for designated high-risk systems, according to guidance published by the European Commission and national competent authorities.
Providers and deployers of systems in listed high-risk categories face documentation, transparency, and risk-management requirements that phase in on a published schedule. Exact deadlines differ by obligation and system type.
Industry groups have asked for clearer technical standards and more time for conformity assessments. Several large cloud and software vendors said they are updating product documentation and customer contracts to reflect the new duties.
Enforcement remains primarily at the national level, with coordination through EU-level bodies. Penalties for serious non-compliance are set out in the regulation; no major fines under the new high-risk chapter had been announced as of our last check.