AI Act
The EU's rules for AI, tiered by how risky the system is.
The EU AI Act is the first comprehensive legal framework for AI. Its structure is risk-based: prohibited uses at the bottom, then high-risk systems with heavy obligations, then transparency rules for things like chatbots, and a mostly-unregulated tier for the rest.
What matters for an organisation outside the EU is that it applies based on where the output is used, not only where the company sits. A model trained in one country and used inside the EU is in scope.
The requirements for high-risk systems are largely about documentation, data governance, human oversight, and logging. That is the same list as a good engineering practice, which is the useful way to read the regulation: compliance and doing it properly overlap more than most people assume.
Related terms
Compliance
Meeting the legal rules that apply to what you built, in the place you built it.
Data Governance
The rules and habits that decide what data exists, who may use it, and how long it lives.
Human in the Loop
A person who reviews, approves, or corrects the agent's work at key points.
Data Privacy
The rules about whose personal data you may hold, and what you must do with it.
AI
A moving label, not a technology. It points at whatever computers have newly learned to do well.
Want this working inside your team?
It starts with one conversation about your workflow.