AI-Risk.ai is live and evolving. Verified evidence is shown where available; deeper coverage, history, watchlists and alerts are being added.
See what's live →Loading current signals and source-level context…
How much of this occupation is AI already touching in observed professional use — and is the broader labor market showing pressure that deserves attention?
The observed-exposure measure combines theoretical AI capability with observed work-related Claude usage. It gives more weight to automation than augmentation, so it is more useful than a pure “could AI do this?” capability score.
Exposure can lead to substitution, augmentation, faster workflows, changed hiring standards or no visible labor-market effect at all. A credible risk view needs market outcomes in addition to exposure.
AI-Risk.ai does not label an occupation “safe” or “doomed” from one dataset. We look for occupation-level divergence in hiring, entry-level access, wages and skill demand before making stronger claims.
Watch whether AI exposure is followed by weaker hiring, fewer entry-level openings, changing wage premiums, new skill requirements or a shift from human substitution toward augmentation. That sequence is far more informative than a single fear score.
Save this occupation as a requested watchlist. This does not activate alerts yet; it tells us exactly which occupation intelligence users want first and puts you on the early-access list for tracking.
Observed-exposure dataset published March 5, 2026.
BLS places this occupation in the very high relative-exposure category after combining three theoretical measures with observed evidence from Anthropic and Microsoft. This gives us an independent federal benchmark instead of relying on one model provider.
It compares occupations with one another. It does not estimate job loss, replacement probability, productivity or future wages.
Official BLS methodology ↗May 2025 OEWS national OEWS estimates add a real outcome layer to the exposure score. Annual changes are useful context, but they are not evidence that AI caused the change.
Prior-year comparison unavailable
Official series ↗Prior-year comparison unavailable
Official series ↗Prior-year comparison unavailable
Official series ↗BLS projections provide a long-run labor-demand baseline prepared from macroeconomic, industry and occupational assumptions. They are not an AI-Risk.ai forecast and do not tell us how much of the projected change will be caused by AI.
-141,800 jobs vs. 2025
Official BLS long-run baseline, not an AI attribution.
Includes growth plus replacement needs; not live vacancies.
These are economy-wide baselines, not occupation-specific causal evidence. They tell us whether the background labor market is strengthening or weakening while AI exposure evolves.
-0.1 pp vs. prior month
Official source ↗-0.02M vs. prior month
Official source ↗-0.09M vs. prior month
Official source ↗AI-related pressure is now part of the official U.S. long-run employment baseline, not only a finding in vendor or academic exposure models. The projection still does not prove that AI alone will cause the decline; BLS incorporates several structural and economic forces.
Office and administrative support workers, employers, workforce planners and training providers — especially where work centers on repetitive information processing, documentation or customer handling.
Occupation-level employment, hiring, wages and openings for customer service representatives, data entry keyers and administrative assistants; annual projection revisions; and the new BLS AI exposure categories.
