Data science extracts insight from data to inform decisions, while AI work builds systems that perceive, predict, or act on their own, and the two overlap heavily in machine learning. The U.S. Bureau of Labor Statistics (BLS) describes data scientists as people who use analytical tools to extract meaningful insights from data, and it counted 275,600 of them in 2025. The Canadian occupation code for data scientists, NOC 21211, lists AI analyst, consultant, and designer among its example titles, so in Canadian job classification the two are not separate occupations.
Key takeaways
- BLS counted 275,600 data scientists and 38,600 computer and information research scientists in 2025, and the second profile is the one that mentions AI.
- BLS projects 35 percent growth for data scientists and 22 percent for research scientists over 2025 to 2035, both much faster than average.
- Canada's NOC 2021 places AI analyst, consultant, and designer titles under the data scientists code 21211.
- Both fields use machine learning, so the practical difference lies in the goal: an explanation or estimate versus a working system.
How the two fields are described in official profiles
The BLS profile for data scientists lists duties such as deciding which data are available and useful, collecting and analyzing data, creating, validating, testing, and updating algorithms and models, using visualization software to present findings, and making business recommendations to stakeholders. Machine learning appears as a tool for classifying or categorizing data. The profile for computer and information research scientists reads differently: it says they design innovative uses for new and existing computing technology, develop algorithms and programming languages, and run experiments that often use data science and machine learning techniques. That profile also names AI as an area where new specialists will be needed.
The BLS numbers side by side
The two profiles are the closest federal matches to the two fields, and the figures below are for 2025 and the 2025 to 2035 projection period.
| Measure | Data scientists | Computer and information research scientists |
|---|---|---|
| Employment, 2025 | 275,600 | 38,600 |
| Median pay, May 2025 | USD 120,230 | USD 140,300 |
| Typical entry education | Bachelor's degree | Master's degree |
| Projected growth, 2025 to 2035 | 35 percent | 22 percent |
| Projected openings per year | About 24,800 | About 2,900 |
The gap in pay and entry education reflects the research orientation of the second occupation, not a ranking of the fields. Projections are estimates that BLS revises.
Canada groups them differently
In the National Occupational Classification (NOC) 2021, version 1.0, code 21211 is titled Data scientists. The NOC profile defines the work as using advanced analytics, including machine learning and predictive modeling, to identify trends, extract information from unstructured data, and generate automated recommendations. Listed duties include applying machine learning, deep learning, and AI techniques, building algorithms such as regression, clustering, decision trees, and neural networks, and developing AI models and algorithms to meet organizational needs. Example titles include artificial intelligence analyst, consultant, and designer. The code usually requires a bachelor's degree in statistics, mathematics, computer science, or a related field, and a master's or doctorate in machine learning, data science, or a related quantitative field is also usually required, along with programming experience.
One dataset, two jobs (illustrative example)
This is an illustrative example, not drawn from either source. A grocery chain has three years of daily sales records. A data science task asks whether last spring's discount raised total sales or only shifted purchases earlier, which needs careful comparison groups and produces a report. An AI task asks a system to place each night's orders for every store automatically, which needs a forecasting model, rules for stock-outs, and monitoring after launch. Both start with the same table and use similar methods. The first ends when people have an answer. The second keeps running, and its errors have costs that someone must watch.
Questions that separate the work in practice
- Is the output a decision-ready finding or software that runs without a person each time?
- Is success measured by the quality of an explanation or by the behavior of a deployed system?
- Who maintains the work after delivery, and how are mistakes detected?
- Does the task need new methods, or is the main effort cleaning data and choosing existing ones?
FAQ
Is data science a part of AI?
Neither official profile describes one field as a subset of the other. Data science draws on statistics and programming, and uses machine learning, which is also a core AI technique. The two share methods without one containing the other.
Do the two fields require different degrees?
BLS lists a bachelor's degree for data scientists and a master's degree for research scientists, and the Canadian code lists statistics, mathematics, and computer science degrees. Individual employers set their own requirements.
Which field pays more?
The BLS median was higher for research scientists in May 2025, but the occupations differ in education and in what they do. Pay for any given role also depends on the employer, the region, and experience.
Are these numbers a forecast of how many AI jobs exist?
No. BLS counts people in defined occupations. Many people who work with AI are classified under other titles, so the counts understate or overstate AI work depending on the definition used.
Related reading
- How AI programs process large datasets
- AI in scientific research
- AI compared with computer science
- AI methods inside data science work
More in AI Basics.
Last verified: October 2026
Sources
- Data Scientists, Occupational Outlook Handbook, U.S. Bureau of Labor Statistics, 2025 data, page updated August 27, 2026
- Computer and Information Research Scientists, Occupational Outlook Handbook, U.S. Bureau of Labor Statistics, 2025 data, page updated August 27, 2026
- 21211 Data scientists, National Occupational Classification 2021 Version 1.0, Employment and Social Development Canada and Statistics Canada