Back to glossary

Automation bias

Otomasyon yanlılığı

D6

Automation bias is over-reliance on automated cues or recommendations that can contribute to omission or commission errors.

Review status: 2027-02-24

Technical explanation

It can affect how people notice, interpret, check, or act on automated output in a decision process.

Conceptual boundaries

Automation bias is not proof that all reliance is irrational or solely an individual-user defect.

Provider-neutral example

A reviewer can accept an automated risk flag without checking the underlying record, or fail to investigate a case because the system did not flag it.

Limitations

The cited studies concern simulated-flight tasks with student and pilot samples and do not establish prevalence or a single cause across all people, tasks, or organizations.

Related concepts

Atomic claims and evidence

  1. 1.1Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source
    Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source role
    Authoritative source
    Exact locator
    Abstract and pp. 231-234
    Supported claim
    Parasuraman and Riley distinguish use, misuse, disuse, and abuse of automation and analyze automation-related errors in human use of automated systems.
    Last verification
    Review due
    Scope limitation
    This foundational human-factors analysis is not a diagnosis of an individual or a claim that every automated recommendation produces bias.
    1.2Regulation (EU) 2024/1689 (Artificial Intelligence Act), consolidated text of 27 July 2026 — D6 human-role slice
    Source
    Regulation (EU) 2024/1689 (Artificial Intelligence Act), consolidated text of 27 July 2026 — D6 human-role slice
    Source role
    Authoritative source
    Exact locator
    Article 14(4)(b), current consolidated text of 27 July 2026
    Supported claim
    Parasuraman and Riley distinguish use, misuse, disuse, and abuse of automation and analyze automation-related errors in human use of automated systems.
    Last verification
    Review due
    Scope limitation
    This foundational human-factors analysis is not a diagnosis of an individual or a claim that every automated recommendation produces bias.
  2. 2.1Regulation (EU) 2024/1689 (Artificial Intelligence Act), consolidated text of 27 July 2026 — D6 human-role slice
    Source
    Regulation (EU) 2024/1689 (Artificial Intelligence Act), consolidated text of 27 July 2026 — D6 human-role slice
    Source role
    Authoritative source
    Exact locator
    Article 14(4)(b), current consolidated text of 27 July 2026
    Supported claim
    The EU AI Act requires assigned overseers of high-risk systems in scope to remain aware of possible automatic reliance or over-reliance, particularly where the system provides information or recommendations for human decisions.
    Last verification
    Review due
    Scope limitation
    This is an instrument-specific oversight obligation; it does not establish the prevalence or cause of reliance in every workplace.
    2.2Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source
    Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source role
    Authoritative source
    Exact locator
    pp. 231-234
    Supported claim
    The EU AI Act requires assigned overseers of high-risk systems in scope to remain aware of possible automatic reliance or over-reliance, particularly where the system provides information or recommendations for human decisions.
    Last verification
    Review due
    Scope limitation
    This is an instrument-specific oversight obligation; it does not establish the prevalence or cause of reliance in every workplace.
  3. 3.1Mosier et al., Automation Bias, Accountability, and Verification Behaviors
    Source
    Mosier et al., Automation Bias, Accountability, and Verification Behaviors
    Source role
    Authoritative source
    Exact locator
    Abstract; pp. 204-208
    Supported claim
    Mosier and colleagues define commission errors as inappropriate action from over-attending to automated information or directives and omission errors as failing to take appropriate action because automated aids do not inform a decision maker of an imminent problem or situation.
    Last verification
    Review due
    Scope limitation
    The article reports two simulated-flight studies with student and commercial-pilot samples; it does not establish prevalence or one cause in every AI-supported decision setting.
    3.2Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source
    Parasuraman and Riley, Humans and Automation: Use, Misuse, Disuse, Abuse
    Source role
    Authoritative source
    Exact locator
    Abstract and pp. 231-234
    Supported claim
    Mosier and colleagues define commission errors as inappropriate action from over-attending to automated information or directives and omission errors as failing to take appropriate action because automated aids do not inform a decision maker of an imminent problem or situation.
    Last verification
    Review due
    Scope limitation
    The article reports two simulated-flight studies with student and commercial-pilot samples; it does not establish prevalence or one cause in every AI-supported decision setting.