AI exposure
Measure of how much an occupation's work could be accelerated by access to AI. In the rubric of Eloundou and colleagues (2023), a task qualifies if a large language model, or software built on one, could cut a person's completion time by at least half; the test of the model alone also requires equivalent quality.
In plain language
The measure asks which work AI might speed up, rather than predicting which jobs will disappear.
In a sentence
- The study estimated AI exposure by examining tasks within each occupation.
- The analyst kept AI exposure separate from a claim that jobs had already disappeared.
Usage note
The measure describes potential economic impact without separating assistance to workers from their displacement. It is not a count of jobs lost.
See also
- Workslop
AI generated work content that passes for good work but lacks the substance to meaningfully advance its task, leaving the recipient to interpret, correct or redo the work.
- Cognitive debt
Metaphorical cost that builds up when people repeatedly rely on AI for thinking they would otherwise do themselves, saving mental effort in the short term while risking later costs to critical inquiry and creativity.
Cite this term.
CC BY 4.0, attribution required.
Orbyt Labs, AI exposure (AI Glossary), Nov 14, 2026. https://
Orbyt Labs Glossaries (Nov 14, 2026), https://
Last updated .