Skip to content
AI Act

Human oversight

In a nutshell

The requirement that trained people can monitor, understand, correct and, if needed, stop a high-risk AI system.

Article 14 requires high-risk systems to be designed so that natural persons can oversee them while in use. The goal is to prevent or minimise risks to health, safety and fundamental rights.

Whoever oversees must be able to: understand the system's capabilities and limits, detect anomalies, be aware of automation bias (the tendency to over-trust the machine), interpret outputs correctly, decide not to use or to override an output, and interrupt the system with a stop button or similar procedure.

The provider designs the system to make this possible; the deployer appoints the people who will do it and ensures they have enough training, authority and support. A "human who signs without looking" does not comply.

Reference: Arts. 14 y 26.2 AI Act

Related terms

What applies to you?

Take the free assessment and find out in a few minutes which obligations your company has and where to start.

Take the free assessment

← Back to the glossaryUpdated: October 2026 · Informational content, not legal advice.