Quick Answer
Artificial intelligence is more likely to change your job than eliminate it outright, according to current US data. The Bureau of Labor Statistics projects 5.9 million more jobs from 2025 to 2035, yet employers cited AI in 120,136 announced US job cuts through September 2026. Risk is highest in clerical, customer service and entry-level office roles.
Key Takeaways
- The Bureau of Labor Statistics projects US employment to grow 3.5%, from 170.3 million to 176.2 million jobs, between 2025 and 2035.
- Office and administrative support jobs are projected to fall 4.0% (752,100 jobs), and BLS links part of that decline to automation tools, including AI.
- Challenger, Gray & Christmas counted 120,136 AI-attributed job cuts through September 2026, about 21% of all announced US cuts.
- Stanford researchers using ADP payroll data found employment for workers aged 22 to 25 in AI-exposed roles 19% below the path of less-exposed peers.
- BLS states that a high AI exposure category is not a forecast of job loss, and the ILO concludes that transformation is the most likely outcome.
Will AI take my job?
Artificial intelligence will probably not take your whole job in the near term, but AI is already reducing demand for some tasks and some entry-level positions. The Bureau of Labor Statistics 2025-35 employment projections, released on August 27, 2026, expect total US employment to grow by 5.9 million jobs, from 170.3 million to 176.2 million. That growth rate of 3.5% is slower than the 10.9% recorded over 2015-25, and BLS attributes much of the projected growth to healthcare and social assistance.
The answer depends heavily on what you do each day rather than on your job title. BLS expects office and administrative support occupations to decline by 4.0%, a loss of 752,100 jobs, and the agency writes that “the continued integration of automation tools, including those powered by AI” is likely to reduce demand for several of those occupations. At the same time, BLS projects computer and mathematical occupations to grow 7.3%, with data scientists growing 34.6%.
This matters because the most useful question is how much of your weekly work consists of tasks that AI tools can already complete without you.
What does the data show about AI job losses so far?
AI job losses so far are real but small relative to the size of the US labor market. The outplacement firm Challenger, Gray & Christmas reported in its September 2026 job cuts report that US employers announced 573,195 job cuts through September, down 39% from 946,426 in the same period of 2025. Employers cited artificial intelligence for 120,136 of those cuts, about 21%, which makes AI the leading stated reason in 2026. In September alone, AI accounted for 3,961 of 43,281 announced cuts.
Those figures need careful interpretation. Challenger counts announced cuts and the reasons companies give, so the data reflects what employers say rather than an independent measurement of whether AI actually performed the eliminated work. TechJournal has covered how large employers have tied 2026 tech layoffs to AI spending, including Meta’s 8,000 layoffs.
Broader labor market studies have found less disruption than the headlines suggest. The Yale Budget Lab, working with Brookings, published an analysis on October 1, 2025 that found the broader labor market had not experienced a discernible disruption in the 33 months after ChatGPT’s release. The researchers reported that the occupational mix was changing only slightly faster than during the arrival of computers and the internet, and that measures of AI exposure showed no clear link to changes in employment or unemployment at that point. The authors also said that better data is needed and that they plan to update the analysis over time.
What is the difference between AI exposure and job replacement?
AI exposure means that AI tools could assist with or complete some of the tasks in an occupation, while job replacement means the occupation itself loses positions. The two are related but are not the same thing. A highly exposed job can still grow if demand for the service rises.
The BLS made this distinction explicit when it published its first AI exposure categories alongside the 2025-35 projections. BLS assigned each of 831 detailed occupations to 1 of 4 relative categories: Low, Moderate, High and Very high. The classification combines 5 external sources. Three are theoretical measures that compare AI capabilities with occupational tasks or abilities, and two are observed measures based on usage data published by Anthropic and Microsoft. BLS states that an exposure category “is not a forecast of employment growth or decline,” and that the categories do not distinguish between automation and augmentation.
International research reaches a similar conclusion. The International Labour Organization’s 2025 refined global index, published on May 20, 2025, estimates that about 1 in 4 workers worldwide are in occupations with some generative AI exposure, rising to 34% of employment in high-income countries. The ILO found that clerical occupations have the highest exposure and concluded that transformation of jobs is the most likely impact. The ILO also found a gender gap: the highest-exposure category covers 9.6% of female employment in high-income countries, compared with 3.5% of male employment.
Which jobs are most and least exposed to AI?
The jobs most exposed to AI are mostly higher-education office and knowledge jobs, while the least exposed jobs involve physical work, hands-on care or emergency response. The BLS employment projections FAQ names web developers, customer service representatives and personal financial advisors as examples of very high exposure. It names maids and housekeeping cleaners, firefighters and dental hygienists as examples of low exposure.
In a February 2025 Monthly Labor Review case study, BLS economists described how they built expected AI effects into the earlier 2023-33 projections. The table below lists the projected change for a selection of those occupations.
| Occupation | What BLS says about AI | Projected change, 2023-33 |
|---|---|---|
| Insurance appraisers, auto damage | AI expected to reduce demand | -9.2% |
| Customer service representatives | Very high exposure; AI expected to reduce demand | -5.0% |
| Medical transcriptionists | AI expected to reduce demand | -4.7% |
| Claims adjusters, examiners and investigators | AI expected to reduce demand | -4.4% |
| Credit analysts | AI expected to reduce demand | -3.9% |
| Paralegals and legal assistants | AI expected to slow growth | +1.2% |
| Lawyers | Growth expected despite AI use | +5.2% |
| Personal financial advisors | Very high exposure; growth still expected | +17.1% |
| Software developers | Growth expected despite AI use | +17.9% |
The table shows why exposure alone is a weak predictor: personal financial advisors have very high exposure, yet BLS projected 17.1% growth. The same BLS article notes that its methods assume technological change follows historical patterns and “are not designed to capture extremely rapid technological change,” so these figures carry real uncertainty if AI adoption accelerates.
Why are entry-level workers feeling the effects first?
Entry-level workers are feeling the effects first because employers appear to be hiring fewer young people into AI-exposed roles rather than laying off experienced staff. The Stanford Digital Economy Lab paper “Canaries in the Coal Mine?” by Erik Brynjolfsson, Bharat Chandar and Ruyu Chen uses ADP payroll data covering millions of US workers through June 2026. In the version revised on August 12, 2026, the authors report that employment for workers aged 22 to 25 in AI-exposed occupations stands 19% below where it would be had it kept pace with less-exposed peers.
The researchers found that the gap operates mainly through reduced hiring rather than increased separations, and that older workers in the same occupations show no comparable decline. The declines are concentrated where AI usage mostly substitutes for human tasks. Where AI mostly complements workers, employment is flat or rising. The authors also report no evidence of widespread, economy-wide job displacement, which is consistent with the Yale Budget Lab findings.
The practical implication is that early-career workers in office, support and junior technical roles face the most immediate pressure, while experience and responsibility for outcomes appear to offer more protection.
How worried are US workers about AI?
US workers are more worried than hopeful about AI at work. A Pew Research Center survey of 5,273 employed US adults, conducted October 7-13, 2024, found that 52% of workers felt worried about future AI use in the workplace and 36% felt hopeful. About 32% said AI will lead to fewer job opportunities for them in the long run, while only 6% expected more opportunities.
The same survey found that only 16% of workers said at least some of their work was done with AI, so many worried workers had little direct experience of what the tools handle well. Usage has since grown: a follow-up Pew survey conducted September 2-8, 2025, found that 21% of US workers said at least some of their work was done with AI. Policy responses are also developing at the state level, as TechJournal reported in its coverage of California’s AI worker protections.
How can you tell whether your own job is at risk?
To tell whether your own job is at risk, assess your daily tasks rather than your job title, then compare them with official occupation data. The following steps give a structured view of your exposure.
- List your tasks. Write down the 10 to 15 tasks that fill a typical week and estimate the share of time each one takes.
- Mark each task by type. Label tasks as routine text or data work (drafting standard emails, summarizing documents, data entry, first-line support), judgment work, relationship work or physical work.
- Look up your occupation. Open the BLS Occupational Outlook Handbook, search your occupation, and read the job outlook figure for 2025-35.
- Check the exposure category. Review the BLS AI exposure categories page to see where your occupation sits relative to the other 831 occupations, and remember that the category is not a job-loss forecast.
- Calculate your routine share. Add up the time spent on routine text or data work. A high share suggests more pressure on your role, especially in an entry-level position.
This exercise cannot predict your employer’s decisions, which also depend on budgets, regulation and how well AI tools perform in your workplace. Treat the result as a guide for skill-building, not a verdict on your career.
What practical steps reduce your risk of losing work to AI?
The practical steps that reduce your risk of losing work to AI involve moving your time toward tasks that AI complements and building evidence that you use AI tools well. The Stanford findings suggest that roles where AI augments workers have held up better than roles where AI substitutes for them.
- Learn the AI tools used in your field. Use the assistants your employer approves, and understand basic concepts such as how AI agents work, so you can judge what the tools do reliably.
- Shift toward judgment and accountability. Volunteer for work that requires decisions, client contact, quality review or sign-off, because those tasks are harder to hand to software.
- Document outcomes, not activity. Record measurable results such as revenue, resolved cases or time saved, which demonstrate value beyond task completion.
- Build adjacent skills. Target growing areas in the BLS projections, such as healthcare, data science and the utilities and computing infrastructure work that BLS links to AI adoption.
- Protect company data. Follow your employer’s AI policy and keep confidential or personal data out of unapproved tools, because a data incident can harm your position more than slow adoption.
Stop and seek professional help if you receive a layoff notice or suspect an AI-related decision was applied unfairly. A state workforce agency can explain unemployment benefits and retraining programs, and an employment lawyer can review whether notice, severance or discrimination rules apply to your case. Public statements from technology leaders vary widely, as shown in coverage of Sam Altman’s comments on AI and jobs, so base career decisions on official data and your own task analysis rather than predictions.
Frequently Asked Questions
Will AI replace my job completely?
AI is unlikely to replace most jobs completely in the near term, according to the BLS and the ILO. Both organizations expect AI to change tasks within jobs more often than it eliminates whole occupations, although some clerical and support roles are projected to shrink.
Which jobs are safest from AI?
The jobs safest from AI involve physical work, hands-on care or emergency response, such as firefighters, dental hygienists and housekeeping cleaners. The BLS lists those occupations as examples of low AI exposure because their requirements do not match well with current AI capabilities.
How many jobs has AI eliminated in 2026?
Employers cited AI in 120,136 announced US job cuts through September 2026, according to Challenger, Gray & Christmas. That count represents about 21% of all announced cuts and reflects reasons employers stated, not an independent measurement.
Are young workers more at risk from AI?
Young workers are more at risk from AI in exposed occupations, according to Stanford research using ADP payroll data. Employment for workers aged 22 to 25 in those roles is 19% below the path of less-exposed peers, mainly because of reduced hiring.
Does high AI exposure mean my job will disappear?
High AI exposure does not mean your job will disappear, because the BLS states that its exposure categories are not forecasts of employment decline. Personal financial advisors, for example, have very high exposure but were projected to grow 17.1% from 2023 to 2033.
