Anthropic study: robots could do 74% of US physical tasks but are cost-competitive on 0.3%
See how this story connectsAnthropic Economics published a robot exposure index built by having Claude classify about 19,000 O*NET task statements and rate the 7,594 physical ones on how structured a setting a robot would need. Today’s machines could perform about 74% of US physical tasks, which make up 34% of working hours, mostly in purpose-built or structured settings; about 1% of work is exposed in unstructured environments. Combined with language-model exposure, about 80% of working time is exposed to one or the other. Nine of the ten most exposed occupations are vehicle operators. Robots are cost-competitive for only 0.3% of tasks; at historical price declines of about 3% a year, reaching 10% would take about 40 years. Exposed workers are more likely to be male and Hispanic, less likely to hold a degree, and earn about $30 an hour less. A backtest from 1977 found earlier robot exposure predicted later wage and employment declines.
Why it matters
The study puts physical automation on the same footing as language models in the work record, and finds cost is the main brake today.
- When
- 30 Sep 2026
- ID
WTA-N-2026-0273- Where
- United States
- Who
- Anthropic; Russell Legate-Yang, Maxim Massenkoff
- Topics
- Jobs and the workforce, AI agents, Research
Summary by wheresthe.ai, from Anthropic. Checked against the source on 12 Oct 2026.
Source copy: download the page as saved on 18 Oct 2026 (PDF). Document the source offers: document 1 (PDF). Also held by the Internet Archive.
Cite this story
WTA-N-2026-0273: “Anthropic study: robots could do 74% of US physical tasks but are cost-competitive on 0.3%.” wheresthe.ai, 30 Sep 2026. https://wheresthe.ai/d/anthropic-robot-exposure-index-2026-09/ (source: Anthropic).
Spotted an error? Email [email protected].