The AI Exposure Index

Powered by EXPO-4 Ultra, our AGI-grade frontier exposure model, trained on 312,448 GPUs and 14.6 quadrillion tokens of human work.

Index updated continuously. Version 4.3.0.

Calculate the AI exposure of

Assessments completed worldwide: 48,213,907

Assessment in progress

Subject:

Overall progress0%

Your AI exposure:

00.000%

Margin of error: ±0.001%
Confidence level: 99.97%

Methodology

The Index applies a proprietary multi-layered exposure framework to the full occupational continuum. Each assessment is computed independently, holistically and, where appropriate, simultaneously.

Our approach integrates quantitative rigour with qualitative depth, ensuring that every result is both precise and meaningful within its own frame of reference.

Full methodological notes are available upon completion of the methodology.

Data sources

The Index draws on a broad and continuously expanding body of primary, secondary and tertiary data, harmonised across all relevant dimensions.

All sources are carefully selected according to criteria established by the sources themselves. Data is refreshed continuously, and occasionally in advance.

Source list available on request. Requests are reviewed on a rolling basis.

Press

The AI Exposure Index welcomes engagement from members of the press seeking to understand the evolving relationship between work, technology and the Index.

Our analysts are available to comment on all developments, including those that have not yet occurred.

Embargoed figures are released at the moment they are calculated.