Automation bias
The tendency of people to rely, or over-rely, on the output of an automated system.
Automation bias is the tendency of people to rely automatically, or to over-rely, on the output of an automated system, including when it is wrong. The EU AI Act names it explicitly, requiring that people assigned to oversee a high-risk AI system be enabled to remain aware of this tendency. For agents, approval screens that people click through by habit are a sign of automation bias, so oversight design has to keep reviews meaningful and rare enough to be real.
Agent Minute explains this term on 13 December 2026.
Related terms
Human oversightMeasures that let people understand, monitor and, when needed, override or stop a high-risk AI system.AI literacyThe skills, knowledge and understanding needed to deploy AI in an informed way and to be aware of its risks.ConfabulationConfidently stated but false or erroneous output from a generative AI system, often called hallucination.Client-Initiated Backchannel Authentication (CIBA)An OpenID flow where an app requests a user's approval, and the user approves on a separate device.