What Is Meaningful Human Control?
Meaningful Human Control is the principle that humans should retain genuine, informed authority over consequential decisions made or supported by AI systems.
Meaningful Human Control — the principle that humans should retain genuine, informed authority over consequential decisions made or supported by AI systems.
Meaningful human control is a higher bar than nominal oversight. It requires that the human has: the competence to evaluate the AI's output, the information needed to make an independent judgment, the authority and time to override, and freedom from automation bias. The concept originated in debates over autonomous weapons but now informs AI governance broadly. The EU AI Act's human-oversight requirement (Article 14) is an attempt to operationalise it for high-risk AI.
Source: EU AI Act, Article 14; UNESCO Recommendation on AI Ethics
Plain-language explanation
Meaningful human control is a higher bar than nominal oversight. It requires that the human has: the competence to evaluate the AI's output, the information needed to make an independent judgment, the authority and time to override, and freedom from automation bias. The concept originated in debates over autonomous weapons but now informs AI governance broadly. The EU AI Act's human-oversight requirement (Article 14) is an attempt to operationalise it for high-risk AI.
See where you stand on AI governance
Take the free 7-question maturity assessment and get a personalised action plan.
Free assessment — 3 minutes →