How were these figures calculated?

The US map covers 106 occupational groups based on the Bureau of Labor Statistics Standard Occupational Classification (SOC) 2018. Together these occupations represent approximately 91 million jobs — roughly 57% of the US civilian workforce (BLS: ~160 million in 2024).

The 106 occupations span all 22 SOC major groups, selected by employment size and sector coverage. Together they represent the occupations most Americans work in and where AI impact is most consequential in terms of aggregate employment.

The BLS SOC 2018 contains 867 detailed occupations. This dataset covers 106 of them across all major sectors. The remaining 761 are not included. Niche specialisations, small occupational categories and occupations below the OEWS reporting threshold are excluded.

The international crosswalk maps each SOC occupation to an ISCO-08 code (International Standard Classification of Occupations), enabling comparison with the Dutch, German, UK and other country editions of this map. Crosswalk source: BLS SOC 2010/2018 to ISCO-08, supplemented with ILO correspondence tables.

Unlike the Dutch map — where AI exposure scores were estimated via LLM and cross-validated with academic indices — the US map uses the Felten, Raj & Seamans (2023) AI Occupational Exposure (AIOE) index as its primary AI score source.

What AIOE measures: The AIOE index maps 10 dimensions of AI capability (as defined by the AI Progress Measurement project at the Electronic Frontier Foundation) to the 52 work abilities defined in O*NET. Each occupation in O*NET has an importance and level rating for each ability. The AIOE score for an occupation is calculated as the weighted overlap between what current AI can do and what the occupation requires.

The raw AIOE scores range from approximately −2.67 to +1.53. For display on this map, scores are linearly normalised to a 1–10 scale using the formula:

A score of 10 indicates the occupation relies most heavily on abilities that AI currently performs well (language, pattern recognition, information retrieval). A score of 1 indicates the occupation relies primarily on abilities where AI has little capability (physical dexterity, spatial reasoning under uncertainty, complex social navigation).

The model answers the question: if current trends continue and AI automation increases, how will employment change by 2030?

It uses the same parametric structure as the Dutch map — combining momentum (recent employment trend) with AI disruption pressure (downward force from automation at higher AI scores). For the US edition, momentum is derived from BLS Employment Projections 2023–2033 growth rates, reverse-engineered into a synthetic historical index.

The US model applies a dampening factor of 0.8 (versus 0.7 for the Netherlands). This reflects the observation that the US labour market tends to respond more quickly to structural trends: at-will employment makes workforce restructuring faster, and there is less regulatory buffering on the employer side.

The AI disruption pressure table for the US also applies stronger pressure at the top end, reflecting the absence of CAO protections and EU AI Act constraints that apply in the Dutch context. The result is presented as a range: conservative (with dampening) to optimistic (without dampening).

The US and Dutch editions share the same parametric forecast structure, but the context in which AI adoption plays out differs substantially.

The AI Exposure Map uses two scoring systems side by side. The primary score on the Dutch map is the Janssen Practical Exposure (JPE), developed for this project. The US map uses the academic Felten AIOE as its primary score, while the Dutch map shows it as an academic reference.

Why both? JPE is used for the treemap, forecasts and career advice (practical decision-making). Felten is shown as an academic reference for transparency and cross-validation.

• This dataset covers 106 of 867 SOC occupations. It represents the largest employers across all major sectors, not the full US labour market. Very small occupational categories and niche specialisations are excluded.
• The AIOE index is based on O*NET ability ratings as of 2023. It does not capture capabilities added by generative AI after that date (GPT-4 Turbo, Claude 3, Gemini 1.5 etc.). Current AI capabilities likely exceed those modelled.
• The AIOE index maps AI to abilities, not tasks. Two occupations with the same AIOE score may differ substantially in how much of their daily work is actually automatable. The score is an approximation.
• Historical indices are synthetic: derived from BLS 2023–2033 growth projections, not independently measured historical time series. This makes the US momentum estimates less precise than the Dutch edition (which uses CBS EBB 2015–2025 data).
• The 2030 forecast is a directional model, not a labour market forecast. It does not account for economic recessions, policy changes, AI winters or breakthrough technologies.
• US-specific vulnerability factors (union density, state-level AI regulation, sector-specific automation barriers) are reflected in the model parameters but not scored per occupation. This is a planned extension.
• Salaries are 2024 medians. Within occupational groups, the spread can be very large — especially in healthcare, legal and technology.
• Not all projected decline is AI-related. Cashiers face self-checkout adoption; office clerks face workflow software. These trends predate generative AI.

License: Creative Commons Attribution 4.0 International (CC BY 4.0)

Cite as: Janssen, S.D. (2026). US AI Exposure Map: 106 occupational groups analysed for AI impact. simondjanssen.nl/en/ai-map/us. Retrieved [date].

AI score source: Felten, E., Raj, M. & Seamans, R. (2023). AIOE index. github.com/AIOE-Data/AIOE

Employment data source: Bureau of Labor Statistics, U.S. Department of Labor. OEWS May 2024.