Open data

Every number on this site, as CSV files anyone can use, including for commercial work, with credit. Estimates, from data checked September 29, 2026; the files are rebuilt whenever the data is.

• Every occupation

  workforce-scan-jobs.csv

  923 rows. O*NET-SOC code, US employment, median pay, the mix of desk, people and body work, the share within reach now and at the end of 2027, 2028 (three paces) and 2030, Anthropic's observed use.

• Every industry

  workforce-scan-industries.csv

  352 rows. NAICS code and level, parent, employment, the share of it task data covers, mean pay, the shares within reach, the value of the hours within reach at the end of 2028, the mix of work.

• Every state, territory, metro and nonmetro area

  workforce-scan-areas.csv

  584 rows. BLS area code, kind, state, and the same columns as the industries.

• How far each has come since GPT-4

  workforce-scan-backcast.csv

  29,744 rows. For every job, industry and area, at each METR frontier measurement since March 2023: the best model measured by then, its four-in-five task length, and the share within reach.

What the numbers mean

Shares are of working time (paid hours for industries and areas), within reach at four in five on the long-run pace unless a column names another pace. Within reach means a model could finish the work at that reliability if the work were set up for it; it is not a forecast of jobs lost or of savings. Value is pay times the cost of an hour (1.43). Blank cells are figures BLS or Anthropic do not publish. How it works has the method, the sources and the limits.

Using it

The files are licensed CC BY 4.0: use them, change them, publish them, with credit. Please credit them as:

Stratus Workforce Scan, Stratus Supply Chain LLC, workforce.stratussc.com, data checked September 29, 2026. CC BY 4.0.

They include information from the O*NET 31.0 Database by the U.S. Department of Labor, Employment and Training Administration (USDOL/ETA), used under the CC BY 4.0 license; O*NET® is a trademark of USDOL/ETA, and Stratus Supply Chain has modified all or some of this information, which USDOL/ETA has not approved, endorsed or tested. Observed use: Anthropic Economic Index, CC BY 4.0. Employment and pay: U.S. Bureau of Labor Statistics. Model ability: METR. Redistributions should keep these notices.

Questions

Write to info@stratussc.com. The same numbers are available to ChatGPT and Claude through the assistant tools.