ADP Research · Oct 21, 2026
ADP Research reports differing employment trends across occupations by AI exposure
Markdown · JSON · Original source
ADP Research published this report on October 21, 2026. The Canaries Dashboard, a project of the Stanford Digital Economy Lab and ADP Research, reports that employment in occupations with high exposure to artificial intelligence shrank year over year in September, while employment in the least- grew. For workers aged 22 to 25, employment contracted overall; high-exposure occupations saw a larger decline, while employment in least-exposed occupations was little changed. The Dashboard reports a similar contrast for workers aged 26 to 30: employment shrank in high-exposure jobs and grew in the least-exposed jobs. For workers aged 22 to 25 in high-exposure jobs, year-over-year contraction had continued for 36 straight months, beginning in October 2023. The figures compare employment trends by age group and occupational exposure; whether AI exposure explains any of the differences remains unknown.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
ADP Research · Article
Source details
Original source date: Oct 21, 2026 (2026-10-21).
Date basis: Date stated on the publisher’s page.
Added to Escape Meridian: 2026-11-04T20:42:00.756Z. First observed: 2026-10-23T21:19:46.415Z. Retrieved: 2026-10-30T06:19:43.502Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Text as retrieved on October 30, 2026; originally published October 21, 2026. The report gives employment trends by but does not establish whether AI caused them.
Open question
What factors, if any, explain the different employment trends across occupations with high and low ?
Source-stated
The Canaries Dashboard reports that employment in occupations with high exposure to artificial intelligence fell year over year in September, while employment in the least-exposed occupations grew.
Employment in occupations with high exposure to artificial intelligence shrank by 0.5 percent in September from a year earlier, while employment in the least-exposed occupations grew by 1.0 percent
Source-stated
The Dashboard reports that employment among workers aged 22 to 25 contracted overall, with a larger decline in high-exposure jobs than in least-exposed jobs.
employment contracting by 2.2 percent year over year. For this age group, employment in jobs with high exposure to AI shrank by 3.7 percent
Source-stated
The Dashboard reports that year-over-year employment contraction for workers aged 22 to 25 in high-exposure jobs had continued for 36 straight months, beginning in October 2023.
marking 36 straight months of year-over-year contraction, a trend that began in October 2023
Source-stated
For workers aged 26 to 30, the Dashboard reports falling employment in high-exposure jobs and growth in the least-exposed jobs.
For workers aged 26 to 30 in jobs with high exposure to AI, employment shrank by 2.8 percent in September from the year prior. For workers of this age group in the least-exposed jobs, employment was up by 0.9 percent.
Escape Meridian interpretation
Escape Meridian reads the reported differences as a comparison of employment trends by age group and occupational exposure, not evidence that AI caused the changes.
according to new data from the Canaries Dashboard, a project of the Stanford Digital Economy Lab and ADP Research.
Source and contributor context
ADP Research · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
AI exposure
In labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Editorial provenance
Version 1. Publication ID: 6421320f-3020-03c7-9bff-10abef092d51.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-04T20:38:38.891Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 3c9d4490d3577df0daae0872dd6381015c97912086e8e6e2064e4dcc38b68f3d. Source text hash: aac9790776c02fdc5a259e97da7befc2810e9e610be5e0eee1eb6edaaa0dcd5b.
Revision and source-check history
Version 1 · 2026-11-04T20:42:00.756Z
ADP Research reports differing employment trends across occupations by AI exposure
ADP Research published this report on October 21, 2026. The Canaries Dashboard, a project of the Stanford Digital Economy Lab and ADP Research, reports that employment in occupations with high exposure to artificial intelligence shrank year over year in September, while employment in the least- grew. For workers aged 22 to 25, employment contracted overall; high-exposure occupations saw a larger decline, while employment in least-exposed occupations was little changed. The Dashboard reports a similar contrast for workers aged 26 to 30: employment shrank in high-exposure jobs and grew in the least-exposed jobs. For workers aged 22 to 25 in high-exposure jobs, year-over-year contraction had continued for 36 straight months, beginning in October 2023. The figures compare employment trends by age group and occupational exposure; whether AI exposure explains any of the differences remains unknown.
Original source
ADP Research · Article
Source details
Original source date: Oct 21, 2026 (2026-10-21).
Date basis: Date stated on the publisher’s page.
Added to Escape Meridian: 2026-11-04T20:42:00.756Z. First observed: 2026-10-23T21:19:46.415Z. Retrieved: 2026-10-30T06:19:43.502Z.
Limits
Text as retrieved on October 30, 2026; originally published October 21, 2026. The report gives employment trends by but does not establish whether AI caused them.
Open question
What factors, if any, explain the different employment trends across occupations with high and low ?
Source-stated
The Canaries Dashboard reports that employment in occupations with high exposure to artificial intelligence fell year over year in September, while employment in the least-exposed occupations grew.
Employment in occupations with high exposure to artificial intelligence shrank by 0.5 percent in September from a year earlier, while employment in the least-exposed occupations grew by 1.0 percent
Source-stated
The Dashboard reports that employment among workers aged 22 to 25 contracted overall, with a larger decline in high-exposure jobs than in least-exposed jobs.
employment contracting by 2.2 percent year over year. For this age group, employment in jobs with high exposure to AI shrank by 3.7 percent
Source-stated
The Dashboard reports that year-over-year employment contraction for workers aged 22 to 25 in high-exposure jobs had continued for 36 straight months, beginning in October 2023.
marking 36 straight months of year-over-year contraction, a trend that began in October 2023
Source-stated
For workers aged 26 to 30, the Dashboard reports falling employment in high-exposure jobs and growth in the least-exposed jobs.
For workers aged 26 to 30 in jobs with high exposure to AI, employment shrank by 2.8 percent in September from the year prior. For workers of this age group in the least-exposed jobs, employment was up by 0.9 percent.
Escape Meridian interpretation
Escape Meridian reads the reported differences as a comparison of employment trends by age group and occupational exposure, not evidence that AI caused the changes.
according to new data from the Canaries Dashboard, a project of the Stanford Digital Economy Lab and ADP Research.
Source and contributor context
ADP Research · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
AI exposure
In labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Editorial provenance
Version 1. Publication ID: 6421320f-3020-03c7-9bff-10abef092d51.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-04T20:38:38.891Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 3c9d4490d3577df0daae0872dd6381015c97912086e8e6e2064e4dcc38b68f3d. Source text hash: aac9790776c02fdc5a259e97da7befc2810e9e610be5e0eee1eb6edaaa0dcd5b.
uncertain · 2026-11-04T20:38:38.891Z · anthropic/claude-opus-5.5
Publication: 6421320f-3020-03c7-9bff-10abef092d51
attribution — pass: The reading correctly attributes the figures to the Canaries Dashboard, a Stanford Digital Economy Lab and ADP Research project, published by ADP Research. The source's claims stay attributed to the Dashboard throughout.
according to new data from the Canaries Dashboard, a project of the Stanford Digital Economy Lab and ADP Research
summary — uncertain: The figures are summarized accurately, and AI exposure is defined in the term library. The reading drops the source's own framing that the 22-25 high-exposure contraction began 12 months after ChatGPT's public release. That is the source's only interpretive link between AI and the trend. The reading then adds that causation is unknown, which flattens the source's implied position. Smallest repair: restore the ChatGPT timing as the source's statement, attributed to the Dashboard. Including the key percentages would also aid readers.
a trend that began in October 2023, 12 months after the initial public release of ChatGPT
date — pass: The reading gives October 21, 2026, at day precision, matching evidence.originalPublished. It is a publisher page, not an archive copy, and no edit flag is set. September data and the October 2023 start are reported as in the source, with no hindsight framing.
support — pass: Each source-stated claim matches its excerpt. The interpretation claim is labeled as Escape Meridian's reading and appropriately avoids asserting causation. The question does not presume an answer. No unsupported benefits or costs are implied.
For workers aged 26 to 30 in jobs with high exposure to AI, employment shrank by 2.8 percent in September from the year prior.
media — pass: No media was selected and no verified candidates were supplied, so nothing is misidentified.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
In labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition referenceIn labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition referenceIn labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition referenceIn labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition referenceIn labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition referenceIn labor economics, how much of a job's tasks an AI system could in principle perform; it is not a forecast of job loss.
Definition reference