Mathematical Modeler: how AI changes this job over time
Our best estimates, shown as ranges and grades - not exact predictions.
Right now, AI can already do about 38% of the day-to-day work in this job, and by 20 years from now that could be around 83%. There are likely to be fewer of these jobs over time - very roughly 40-75% of the 2024 number, 20 years out. Getting your first job here is fairly easy today, and it looks set to get harder. This job has no special protection, so the trick is to keep building skills AI cannot copy.
What we assume: AI keeps getting cheaper and better; robots arrive more slowly - small effect by ~2031, bigger by ~2036, widespread by the mid-2040s. "Number of jobs" means how many jobs there will be compared with 2024 (100% = the same). "How hard to get in" runs from A (easy) to E (very hard).
How a Mathematical Modeler job changes over time
AI is already doing some of the everyday work in this kind of job, like writing, sorting and simple admin. The job is not going away, but it is changing - and because there is no licence or special rule protecting it, it pays to watch how fast the tools get better. The people who do best will learn to use AI well and head for the parts that still need a real person: good judgement, working with others, and being trusted to be fair.
AI will do more of the simple, repeating tasks, and employers will expect you to use it. The job gets more interesting for people who adapt - and harder to get into if all you offer is the routine work.
More of the middle of the job is done by AI. People who moved towards judgement, looking after clients, and skills from more than one area do well. The pure tick-box version of the job gets much thinner.
Nobody can really see this far ahead. With no special protection, how this job ends up depends on how clever AI gets - which no one knows. Build skills you can take anywhere, and expect to learn new things more than once.
The honest bottom line: this job is changing, not vanishing. The routine work goes to AI, and there are likely to be fewer of these roles over time. What lasts is the human side - making fair decisions, sorting out problems, and being trusted. Aim for that, learn to use AI well, and you will be in a strong place.
How to aim for a Mathematical Modeler career
You're looking ahead at this job. By the time you join, AI will already do more of it - so aim for the part that will still need a person.
Sources: exposure dial - Anthropic labour market research (2026), observed real-world AI usage by occupation. Job-security category and forecast - OpenAI, "The AI Jobs Transition Framework" (Richmond, 2026, OpenAI Economic Research), CC BY 4.0, matched to "Statisticians" (15-2041.00). Scorecard grades and verdicts are CourseMap editorial judgment - we show forecasts as forecasts and own our conclusions.