The European Union's regulation for AI, the first broad AI law, which sorts AI systems by risk and sets obligations accordingly.
It groups systems into tiers, from banned uses to high-risk systems that carry the heaviest obligations, down to limited and minimal risk. High-risk systems face requirements around risk management, data quality, transparency, human oversight, and record keeping. Many customer-facing and decision-making agents fall into scope, which makes ongoing oversight and evidence a practical requirement, not just a legal one.
Does the EU AI Act apply to AI agents?
Yes, where an agent falls into a regulated use. The Act regulates by how a system is used and the risk it carries, so a customer-facing or decision-making agent can sit in the high-risk tier and carry those obligations.
Related: AI Compliance, Model Risk Management, AI Agent Governance
See also: AI Agent Drift in Regulated Environments