Automation bias

We find that AI use impairs conceptual understanding, code reading, and debugging abilities, without delivering significant efficiency gains on average.

An RCT with junior developers: the AI-assisted group never hit and resolved errors, so never built the debugging and code-reading skill needed to check outputs later — direct experimental evidence for the oversight-competence problem, not just the labor-market one.

This database tracks legal decisions where the use of AI... is addressed in more than a passing reference by the court or tribunal.

The live tracker of court decisions involving hallucinated AI content reached 1,822 cases by August 1, 2026 — the largest running dataset of trained professionals filing unchecked AI output under sanction risk, i.e. automation bias measured in the wild.

people can perform at a pretty high level, because they're basically borrowing skills from the AI, but are not developing those skills themselves.

Nature's synthesis of the first wave of deskilling evidence — the endoscopy adenoma-detection decline, the coding skill-formation RCT, pre-AI precedents — marking the point where 'AI erodes the human backup' moved to a measured result across domains.

users with high overreliance show frequent copy-paste, skipping initial comprehension, repeated LLM references, coarse locating, and accepting misinformation despite hesitation.

Moves overreliance research from self-report to observable interaction behavior, identifying signatures that predict accepting LLM misinformation — a path to detecting in real time when users have stopped checking outputs.