What Is Automation Bias?
Automation Bias is the tendency for humans to over-rely on automated AI recommendations, deferring to them even when their own judgment or other information suggests a different conclusion.
Automation Bias — the tendency for humans to over-rely on automated AI recommendations, deferring to them even when their own judgment or other information suggests a different conclusion.
Automation bias is one of the most significant human oversight failures in AI deployment. Studies in aviation, radiology, and clinical decision support consistently show that when an AI system provides a recommendation, humans are likely to accept it without critical evaluation — even when the AI is demonstrably wrong. The EU AI Act's human oversight requirements (Article 14) are designed to counter automation bias by ensuring deployers actually review AI outputs rather than rubber-stamp them. Designing effective human oversight requires understanding automation bias and counteracting it in interface and workflow design.
Source: Skitka et al. (1999); EU AI Act, Article 14
Plain-language explanation
Automation bias is one of the most significant human oversight failures in AI deployment. Studies in aviation, radiology, and clinical decision support consistently show that when an AI system provides a recommendation, humans are likely to accept it without critical evaluation — even when the AI is demonstrably wrong. The EU AI Act's human oversight requirements (Article 14) are designed to counter automation bias by ensuring deployers actually review AI outputs rather than rubber-stamp them. Designing effective human oversight requires understanding automation bias and counteracting it in interface and workflow design.
See where you stand on AI governance
Take the free 7-question maturity assessment and get a personalised action plan.
Free assessment — 3 minutes →