episcope

Definition

Deferring to an automated output even when it is visibly wrong. It takes two forms: acting on a recommendation that is wrong, and failing to act because the system raised no flag. Both were measured on flight decks before AI meant a chatbot.

What is at stake

The oldest name here for the newest problem. Thirty years of cockpit and control-room studies describe what the 2026 experiments are finding again, which means the field has a design literature already, and its answer was never that people should try harder.

Where the word came from

  1. The ironies of automationLisanne Bainbridge, 1983
  2. Automation biasMosier and Skitka, 1990s, errors of commission and omission on the flight deck
  3. The crossover pointParasuraman and Manzey, 2010, the accuracy level above which reliance helps more than it hurts
  4. Falling asleep at the wheelDell'Acqua, 2022, the same bias on a hiring task

How episcope came to it

Kathleen Mosier and Linda Skitka, in the human factors literature on flight-deck automation, mid 1990s. Their 1999 study with Mark Burdick, peer reviewed in the International Journal of Human-Computer Studies, put numbers on both forms of the bias in a simulated cockpit. The term sits between Lisanne Bainbridge's 1983 paper on one side and Fabrizio Dell'Acqua's 2022 recruiter experiment on the other, as the parent of both.

What to watch for

The counterweight travels with the term. Raja Parasuraman and Dietrich Manzey, in a peer-reviewed 2010 review in Human Factors, put the crossover at roughly 70 percent accuracy under high workload. Above it, relying on the system helps more than it hurts, so reliance is a cost only below the line. The figure comes from cockpits and control rooms rather than from generative AI, and it measures the task outcome, not the operator's skill six months later.

Papers and articles behind it

Skitka, L. J., Mosier, K. L., & Burdick, M. (1999). Does automation bias decision-making?. International Journal of Human-Computer Studies, 51(5), 991-1006.

Peer reviewed

Parasuraman, R., & Manzey, D. H. (2010). Complacency and Bias in Human Use of Automation: An Attentional Integration. Human Factors, 52(3), 381-410.

Peer reviewed

Read next

Back to the glossary

Curious?

The glossary is free and it stays that way. If you want to work through what it means for your own team, twenty minutes is usually enough to tell whether there is something here for you.