Feb 2026 · 178 Countries · 923 Occupations

Mapping AI adoption across the global economy

DOI: 10.5281/zenodo.20320112 CC BY 4.0

GAIA provides two distinct occupation-level AI exposure scores — GAIA-E (generative-AI era, Eloundou et al. 2024) and GAIA-B (supervised-ML era, Brynjolfsson et al. 2018) — plus Anthropic's observed Claude.ai behavioral data across 178 countries. The two scores capture different technology paradigms and are kept separate by design.

Explore Occupations ↗ Country Data →

Data access by request. Cite as: Aghabarari (2026) DOI: 10.5281/zenodo.20320112

178
Countries tracked
via Claude.ai behavioral data
923
Occupations rated
O*NET-SOC standard codes
70%
Global task-success rate
across all Claude.ai requests
34.5%
Avg AI exposure
across all occupations (beta)
How AI is being used

Work is the leading use case — ahead of personal and coursework

Nearly half of all Claude.ai requests in February 2026 were work-related. Personal use follows at 42%, and educational use accounts for 12% — signaling that AI has already crossed from experimentation into daily professional workflows.

  • 45.2%
    Work
  • 42.3%
    Personal
  • 12.4%
    Coursework
Occupation exposure

Which jobs are most — and least — exposed to AI?

GAIA-E (Eloundou et al. 2024) estimates the share of tasks performable by GPT-4 with software tools — the generative-AI era measure. Knowledge-intensive roles lead; physical and service roles trail. See also GAIA-B (Brynjolfsson et al. 2018) for the pre-generative-AI supervised-ML baseline.

Highest AI exposure
Mathematicians
100%
Proofreaders & Copy Markers
98%
Blockchain Engineers
97%
Correspondence Clerks
96%
Court Reporters
96%
Lowest AI exposure
Athletes & Sports Competitors
0%
Orderlies
0%
Cooks, Fast Food
0%
Cooks, Short Order
0%
Dining Room Attendants
0%
Browse all 923 occupations →
Human–AI collaboration

How people collaborate with AI at work

Claude.ai logs show six distinct collaboration modes. Directive use — giving direct instructions to complete tasks — dominates, followed by task iteration and learning.

32.6%
Directive
User issues explicit instructions for AI to execute
25.6%
Task Iteration
Refining outputs through back-and-forth exchanges
22.4%
Learning
Using AI to build understanding or gain knowledge
11.5%
Feedback Loop
Giving AI structured feedback to improve responses
4.9%
Validation
AI checks or confirms human-generated work
3.0%
None / Other
Unclassified or minimal interaction patterns
Research

Papers behind the index

Working papers on measuring AI exposure, testing whether theoretical exposure predicts real capability, and tracing AI's effects through credit markets and employment.

Working Paper · 2026

GAIA: A Global AI Adoption Index

Observed usage, theoretical exposure, and pre-GPT baselines across 923 occupations and 138 countries.

Read paper → Preliminary · May 2026

Exposure Without Performance

Does theoretical AI exposure predict actual AI capability? Evidence from the United States — and a surprising negative result.

Read paper → Forthcoming · 2026

AI Exposure, Credit Markets & Employment in Brazil

How occupational AI exposure transmits through bank credit into employment and wages, with Banco Central do Brasil.

Read paper →
View all research →