What is Article 14 of the EU AI Act?
Article 14 of the EU AI Act is the human oversight provision. It requires that high-risk AI systems be designed and developed so that natural persons can effectively oversee them while they are in use. That is the whole idea in one sentence: if an AI system can make or shape decisions that affect people's health, safety, or rights, a human has to be able to understand it, step in, and stop it, by design. Primary source: EUR-Lex, Regulation (EU) 2024/1689, Article 14.
What does the text actually say?
That high-risk systems shall be designed, including with appropriate human-machine interface tools, so they can be effectively overseen by natural persons during the period in which they are in use, in order to prevent or minimise risks to health, safety, and fundamental rights. The emphasis on effectively is the point: presence is not the standard, the ability to intervene is. See the European Commission AI Act Service Desk for the article-level reference.
Which systems does it cover?
Article 14 applies to high-risk AI systems, the category the Act defines for AI used in sensitive areas such as employment, credit, essential services, law enforcement, and more. If a system is high-risk, the oversight obligation attaches. Whether a given system is high-risk is its own question, set out elsewhere in the Act.
Why it is written this way
Because the drafters had seen the failure mode this whole field is about: a machine decides, a person is affected, and no one can say who could have stopped it. Article 14 is the law's answer, oversight that is built in rather than promised, the legal echo of what we call human judgment infrastructure. For the deeper argument about why nominal oversight so often fails, see Human Oversight Is Mostly Theater.
Read next
- What does meaningful human oversight mean?
- The Article 14 oversight measures, explained
- Back to the Article 14 hub
Sources
This page explains the law and is not legal advice.