What this occupation does
Data scientists develop and implement a set of techniques or analytics applications to transform raw data into meaningful information using data-oriented programming languages and visualisation software. The role spans data preparation, modelling, statistical analysis, dashboarding, and stakeholder communication.
The exposure score in context
The ILO 2025 refined Generative AI Occupational Exposure Index places data scientists in the moderate exposure gradient. ILO 2025 places data scientists in the moderate exposure gradient. Routine data wrangling and standard reporting are highly exposed; modelling, feature engineering, and stakeholder interpretation are augmentation-prone but not displaceable at task level.
The mapping uses ISCO-08 code 2120 (BLS-published SOC-to-ISCO crosswalk). The full methodology, including the dominant-match rule for one-to-many crosswalks, is at /methodology/#algorithm.
The top five tasks, classified
The top five O*NET 30.2 tasks for this occupation, each tagged Displaceable / Changing / Growing per the Brookings 2024 task-level rubric. The tag definitions are at /glossary/#displaceable-task, /glossary/#changing-task, and /glossary/#growing-task.
- Changing: Apply feature selection algorithms to models predicting outcomes of interest, such as sales, attrition, and healthcare use. AutoML and AI-augmented feature selection are widely used; final modelling judgement remains human.
- Growing: Apply sampling techniques to determine groups to be surveyed or use complete enumeration bases. Sampling design involves statistical judgement that grows in importance as AI handles more execution.
- Changing: Compare models using statistical performance metrics, such as loss functions or proportion of explained variance. Model comparison is heavily AI-augmented; deployment judgement remains human.
- Displaceable: Develop and implement procedures for cleaning data and handling outliers. Standard data-cleaning is technically and contextually feasible for current generative AI.
- Growing: Present analytical findings to non-technical audiences. Stakeholder communication is augmentation-prone per Brookings 2024 and grows with the volume of analyses.
What is growing in this role
The BLS Employment Projections 2024-2034 outlook for data scientists is much faster than average (+33.5% projected change, +82.5k jobs). Source: BLS Employment Projections 2024-2034.
Per the WEF Future of Jobs Report 2025, the top three growing skills relevant to this role are: AI and big data, Analytical thinking, Networks and cybersecurity. The skills are mapped to the occupation's O*NET skills profile.
Brookings 2024 finds data-science tasks across the spectrum: data preparation is exposed; modelling judgement, feature engineering, and stakeholder communication are augmentation-prone.
Data Scientists in the WEF Future of Jobs Report 2025
The WEF Future of Jobs Report 2025 ranks big data specialists as its single fastest-growing role by proportional growth to 2030; the closely related data-analyst work tracks the same demand surge as organisations scale their use of data and AI. Source: World Economic Forum, Future of Jobs Report 2025 (January 2025). The WEF lists are occupation-family level and global, ranked by change to 2030; they are distinct from the BLS 2024-2034 US projection above. The full named growing and declining lists are on what is growing.
Similar occupations
O*NET 30.2 lists the following related roles. Each links to its own deep dive where one is published.
- Software Developers
- Financial Analysts
- Marketing Managers
- Computer and Information Research Scientists
- Accountants and Auditors
Industry context
This role sits primarily in the Technology industry. The industry-level rollup includes the cross-occupation exposure profile and the BLS-published industry-level outlook.
How this assessment was made
The full methodology is at /methodology/: ILO 2025 refined index for the gradient, Brookings 2024 rubric for the task tags, BLS 2024-2034 for the growth outlook, WEF 2025 for the skills demand. The pre-empted critiques are at /how-to-argue-with-this/.