Will AI Replace My Job?

Probably not, but it will change a lot of jobs, and the federal data shows which ones. The Bureau of Labor Statistics sorts 786 occupations into four relative AI exposure categories: low, moderate, high and very high. Exposure means AI could be used to help with or complete some of the work. It does not mean the job goes away, and BLS says plainly that exposure does not imply job loss, productivity gains, automation probability or wage effects.

That is why the tool pairs the category with the BLS employment projection for 2025–2035. Of the occupations in the high and very high categories that BLS projects, 295 of 389 (76%) are expected to hold or add jobs. A highly exposed job can still grow when AI makes the people in it more productive and demand for the work rises.

Low exposure (26%)

203

Moderate exposure (25%)

194

High exposure (24%)

190

Very high exposure (25%)

199

Occupations per BLS AI exposure category, one per SOC code, from the BLS Employment Projections supplemental table as held in the NueCareer career bank.

The work AI reaches first is routine information work: processing records, working at a screen, repeating the same tasks. The table shows the three largest occupations in each category by employment, with the BLS outlook beside them. Retail salespersons are the largest occupation in the very high category, and warehouse workers the largest in the low one.

Workers: BLS OEWS national employment, May 2025. Projected change: BLS Employment Projections, 2025–2035; all occupations 3.5%.

Read across a row and the verdict follows from the two numbers. For retail salespersons the tool says: "This is one of the most exposed jobs, and BLS projects it to shrink. Exposure is not a forecast of job loss on its own, but here the two point the same way." The same logic runs for every occupation in the picker.

The short answer is built from the two federal numbers only, never from our own score. The BLS category sets how much AI touches the work; the BLS projection says whether employment is expected to grow or fall. Each combination has its own answer, worded to say no more than the data can support.

• Low exposure. Very unlikely. These jobs are among the occupations least exposed to current AI, and most of the work is the kind current AI tools cannot do.
• Moderate exposure. Unlikely. If BLS projects the job to shrink, the answer says so, and that the cause is not mainly AI.
• High exposure. Not replaced but changed when the job is projected to grow; exposed and shrinking when it is not.
• Very high exposure. Reshaped more than most when the job is still growing; one of the most exposed and shrinking jobs when both numbers point the same way.

Under the answer the tool shows the job's projected change against the 3.5% average for all occupations, its annual openings and its median pay, so you can see the size of the job market, not just its direction.

The BLS category says how exposed an occupation is, not why. For the why, the tool reads O*NET, the federal database of what each job involves, which rates how important every work activity is and how often each working condition comes up. Activities such as processing information, working with computers and documenting records overlap most with what AI handles well. Face-to-face discussion, hands-on work and building relationships hold up best.

NueCareer combines those ratings into an AI-resilience score from 0 to 100, and the tool shows the three ratings that push it up and the three that pull it down. The score is our analysis, not a federal figure, and the methodology page explains how it is built. It is there to explain a job's exposure, not to overrule it.

High exposure is a reason to plan, not to panic. Three moves work for most people: get good at the AI tools that touch your work, since the people who use them well are the ones employers keep; lean into the parts of the job that hold up, usually judgment, client contact and hands-on work; and know which nearby careers are less exposed. The tool lists up to three related careers in a lower BLS category for each job, drawn from O*NET's related occupations.

281 of the largest and most searched occupations have a full page of their own, with the tasks O*NET lists and more alternatives. Browse the jobs most exposed to AI and the jobs least exposed to AI, or take the career quiz to find work that fits you.

• It rates occupations, not you. The category describes the typical mix of work in an occupation, not your employer or your role.
• It is relative. The four categories rank occupations against each other. Low does not mean zero and very high does not mean certain.
• It is a snapshot of AI ability. The category reflects AI as the underlying studies measured it. Capabilities move faster than the federal data.
• Projections are not promises. The BLS outlook for 2025–2035 is the best federal estimate of employment change, and it is revised every year.
• Specialties share a parent's data. BLS does not rate specialty careers on their own, so the tool shows the parent occupation's figures and says which one.