Adaptation · Labour & consumers
Exposed vs unexposed occupation employment gap (Revelio)
Employment in the most AI-exposed occupations relative to the least exposed, since October 2022, from Revelio Labs' AI Labor Market Tracker (online profile data).
Reason through thisThe reading
The gap widened from ~4% to ~6% in a month of releases; direction matters more than level here.
1 reading drawn from revelio.us_exposed.growth_gap.m; 1 observation in the evidence.
Source: Revelio Labs AI Labor Market Tracker (US). Revelio Labs, AI Labor Market Tracker.
The numbers
| as of | reading | value | record |
|---|---|---|---|
| 2026-07-31 | us exposed | -6.0% | reported |
Why it matters
An independent, differently sourced read on the same question as the Canaries payroll series.
What cuts against it
Profile-based data over-represents white-collar workers; the event study's control group is not immune to AI; single source.
How it is measured
- series
revelio.us_exposed.growth_gap.m - source Revelio Labs AI Labor Market Tracker (US) · default tier 6 · Revelio Labs; short quotation
- proxy deployment, welfare · cadence monthly · valve adoption to adaptation
Normal = a relative gap of at most 5 points over nearly four years, within the drift between occupation groups in any period; fast = a 10-point or larger gap. Between is `emerging`.
Applied to revelio.us_exposed.growth_gap.m.
Latest readings
- 2026-07-31 us_exposed · -6.0%as of 2026-07-31
Why this status, and its history
Revelio Labs AI Labor Market Tracker, August 2026 edition (data to July 2026): employment in the most AI-exposed occupations is down ~6% relative to the least exposed since pre-ChatGPT, widened from ~4% in the previous edition. Between the normal band (>= -5%) and the fast band (<= -10%). Single source. Initial seed.
- 2026-09-10unmeasured to emergingconf — → 65 · initial seed (model-drafted)
Revelio Labs AI Labor Market Tracker, August 2026 edition (data to July 2026): employment in the most AI-exposed occupations is down ~6% relative to the least exposed since pre-ChatGPT, widened from ~4% in the previous edition. Between the normal band (>= -5%) and the fast band (<= -10%). Single source. Initial seed.
source1
Why this confidence
Rests on 1 observation from Revelio Labs AI Labor Market Tracker (US); best evidence tier 6, grade B; stale since 2026-07-31.
Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; 0–49 limited or vague evidence.