EU AI Act

The EU AI Act (Regulation (EU) 2024/1689) is the European Union's horizontal regulation for AI, classifying systems by risk and imposing obligations on providers, deployers, importers, and distributors.

What is EU AI Act?

The EU AI Act categorises AI systems as prohibited, high-risk, limited-risk, or minimal-risk. High-risk systems (including AI in employment, credit, education, biometric identification, critical infrastructure, and certain product-safety contexts) carry obligations on data quality, technical documentation, transparency, human oversight, accuracy, robustness, and cybersecurity. The Act also creates rules for general-purpose AI models, including transparency and copyright duties. Phased application runs from February 2025 (prohibitions) through August 2026 (most high-risk obligations) into 2027.

How does EU AI Act apply to enterprise AI?

Any enterprise selling into the EU, processing EU data subjects, or deploying AI for EU users falls within scope, regardless of headquarters location. Compliance work involves classification, conformity assessment, documentation, post-market monitoring, and incident reporting.