Will AI Cause Job Losses? Key Findings From BLS 2025 Data

Will AI Cause Job Losses? Key Findings From BLS 2025 Data

The federal government's own labor economists are not predicting an employment collapse from AI. The Bureau of Labor Statistics treats AI the same way it has treated every prior technology shift: as a gradual, structural force that reshapes hiring demand across specific occupations rather than eliminating employment broadly, according to a 2025 Monthly Labor Review article on how BLS incorporates AI into its employment projections.

That framing matters for anyone trying to evaluate their own career exposure. Will AI cause job losses? The honest answer depends entirely on which occupation you're asking about. BLS projects declining employment in several roles with heavy AI exposure while simultaneously projecting above-average growth in others, the article explains. The divergence runs occupation by occupation, not economy-wide.

Here's what that means for workers, students, and career changers trying to make sense of the data.

What federal employment projections can and cannot tell you about AI job displacement

BLS is clear on a point that most coverage skips: its projections measure AI's employment impact, not its broader economic or social reach, the Monthly Labor Review article notes. That distinction changes how you should read any growth or decline figure. An occupation can be significantly reshaped by AI, with tools handling more of the underlying work, while total employment in that field still grows if overall demand for the service keeps rising.

The projection methodology also has a built-in assumption worth understanding. BLS methods are designed to capture structural, gradual technological change, and the agency explicitly acknowledges those methods are not built to model extremely rapid adoption, per the same source. If AI adoption accelerates beyond historical norms, the current projections would undercount disruption.

Three points help prevent misreading what the figures actually show. A projected decline reflects lower total employment accumulated over the full 2023-33 period, not a prediction of mass layoffs at a specific moment. Even occupations projected to shrink still generate job openings each year as workers retire or move on. And national projections do not reflect local labor markets, which can diverge significantly from the U.S. aggregate.

These projections are a credible starting point for career planning, not a complete answer. They indicate the direction of hiring demand. They say nothing about wages, job quality, regional availability, or how quickly change will arrive at your specific employer or in your industry.

Will AI cause job losses in every occupation? The declining-demand picture

Over the 2023-33 period, AI is expected to most directly affect occupations whose core tasks are most easily replicated by generative AI in its current form, according to BLS. Medical transcriptionists are projected to decline 4.7% and customer service representatives 5.0% through 2033. Claims adjusters, examiners, and investigators face a 4.4% projected decline; auto damage insurance appraisers face a steeper 9.2% drop. Credit analysts are projected to decline 3.9% as AI tools absorb more of the analytical and processing work central to the role, the projections show.

High AI exposure and declining employment demand are not the same thing, though. The legal profession illustrates why. A 2023 study cited by BLS estimated that roughly 44% of legal tasks are susceptible to automation, ranking the profession among the most AI-exposed fields. Yet lawyer employment is projected to grow 5.2% through 2033, roughly in line with the 4.0% average for all occupations, because continued demand for legal services offsets what AI handles on the back end, BLS projects. Task exposure is one input; total demand for the service is the other. Both shape the employment outcome.

For workers in roles facing projected declines, the practical question is not whether the projection applies to you personally but what it signals about where hiring demand is heading. Identify which parts of your current role require human judgment, client trust, or regulatory accountability. Those are the skills most worth documenting on a resume and building toward in adjacent roles. A sample of current job postings in your field can also reveal whether employers are already adding AI-tool proficiency or automation oversight as qualifications, which often signals a shift in hiring criteria before it becomes standard.

Where the AI employment impact runs in the opposite direction

Some of the fastest-growing occupations through 2033 are directly tied to AI infrastructure. Software developers are projected to grow 17.9%, more than four times the 4.0% average for all occupations, per BLS. Database administrators are projected to grow 8.2% and database architects 10.8% over the same period, the data shows. These roles expand in part because AI systems require substantially more data infrastructure to build, train, and manage.

A second cluster of technical roles is growing faster than average for reasons that include but extend beyond AI. Electrical and electronics engineers are projected at 9.1% growth, computer hardware engineers at 7.2%, aerospace engineers at 6.0%, and aerospace engineering technologists and technicians at 7.9%, according to BLS. These occupations benefit from AI-enabled design tools, but their growth also reflects demand from defense, energy, and infrastructure sectors independent of AI.

A third pattern is worth separating out: some fields are growing despite significant AI exposure, not because of it. Personal financial advisors are projected to grow 17.1% through 2033, and civil engineers 6.5%, the projections show. In both cases, rising demand for the service drives the projection. AI may handle more of the underlying analytical work while the human professional remains the accountable party.

A high projected growth rate is a useful signal, not an employment guarantee. Entry into technical fields typically requires specific credentials that vary by employer and state. A 17.9% national growth projection for software developers does not tell you whether roles are available in your market, what they pay at entry level, or which skills employers are currently screening for versus simply preferring.

How to evaluate your own AI jobs outlook by career stage

Workers in occupations projected to decline should start with two concrete steps. Look up your role in the BLS Occupational Outlook Handbook and note both the projected growth rate and the projected annual openings, which BLS reports separately. Then review a sample of current postings for your role on major job boards and watch for language shifts: new references to AI tool proficiency, automation oversight, or data governance often indicate where employers are already redirecting the role. Skills that tend to transfer most cleanly into adjacent growing roles include judgment-based assessment, client communication, and regulatory or compliance knowledge.

Students selecting a major or early career path should avoid treating a single growth projection as a career decision. Cross-reference the Handbook entry for your target role against current local postings to confirm that demand actually exists in your region. Check whether entry-level roles require a specific degree or whether employers are also hiring candidates with certifications or equivalent experience. Compare projected growth percentage against total projected annual job openings; a faster-growing but small occupation may add fewer total positions than a slower-growing but large one.

Career changers considering a technical pivot face the most layered decision. Before targeting software development, database management, or engineering roles, identify the specific credential or experience gap between your current background and what employers are actually screening for. Many technical postings list preferred qualifications, specific cloud platforms, programming languages, security certifications, that are not universally required but appear consistently enough to function as practical filters. Before enrolling in any training program or bootcamp, verify that employers in your target field recognize that credential.

One comparison every reader should make regardless of career stage: projected employment growth and projected annual job openings are different figures, and BLS reports both in the Occupational Outlook Handbook. An occupation with declining total employment can still generate substantial annual openings as current workers retire or leave. Checking both figures gives a more accurate picture of actual hiring volume than growth rate alone.

What the evidence supports and where it stops

BLS projects AI as a reorganizing force across occupations, not a broad employment collapse. Some roles face sustained hiring-demand pressure through 2033; others are growing well above average; many are holding steady despite significant task-level AI exposure, the Monthly Labor Review article concludes.

The framework has a ceiling. It assumes AI adoption proceeds at a pace consistent with past technology cycles. If that assumption proves wrong and adoption accelerates sharply, the projections would understate disruption, and adjustment costs would fall hardest on workers in the roles with the least flexibility to adapt, as BLS acknowledges.

National projections tell you the general direction of demand. Your actual employment situation depends on your specific occupation, employer, region, and credential profile. A declining national projection is not a fixed outcome; a growing one is not a promise.

Go to the BLS Occupational Outlook Handbook and look up your current or target occupation. Note the projected growth rate and the projected annual openings, then pull a sample of current postings for that role in your target location. The gap between what the projections describe and what employers are asking for right now is where your preparation should focus.