BLS projects AI-era shifts in office and tech jobs
New federal employment projections point to a changing mix of U.S. jobs over the next decade: demand may weaken in several repetitive office-support occupations while technical roles tied to software, data and cybersecurity expand.
The Bureau of Labor Statistics released its 2025-35 employment projections on August 27, 2026. BLS projects total U.S. employment will rise from 170.3 million jobs in 2025 to 176.2 million in 2035, an increase of 5.9 million jobs, or 3.5%, under the agency’s assumptions about labor supply, productivity, industry output and full employment.
Alongside the release, BLS introduced a data product comparing occupations by their theoretical and observed exposure to artificial intelligence. The agency says the product is meant to help users compare occupations; it is not a measure of which jobs will necessarily disappear.
Office-support work has the largest projected decline
Among the 22 major occupational groups, office and administrative-support work is projected to decline the fastest from 2025 to 2035. BLS projects a 4.0% decrease, equal to about 752,100 fewer jobs.
BLS links the expected decline to the continued integration of automation tools, including AI-powered systems, into workplace workflows. Those tools can automate repetitive tasks, improve efficiency and allow employers to handle some processes with fewer labor hours. The projection also reflects broader changes in industry demand, work organization and technology, not an isolated estimate of AI’s effect.
In a separate BLS comparison covering 2024-34, employment of customer service representatives is projected to fall 5.5%, or 153,700 jobs. Legal secretaries and administrative assistants are projected to decline 5.8%, or 9,000 jobs. Credit authorizers, checkers and clerks are projected to fall 6.2%, or 700 jobs, while procurement clerks are projected to decline 8.7%, or 5,400 jobs.
Technical occupations show stronger projected growth
The detailed BLS table also shows stronger projected demand in several technology-related occupations during 2024-34. Data scientists are projected to grow 33.5%, adding 82,500 jobs. Information-security analysts are projected to grow 28.5%, adding 52,100 jobs.
Computer and information research scientists are projected to grow 19.7%, or 7,900 jobs, and software developers are projected to grow 15.8%, or 267,700 jobs. The software-developer figure is the largest numerical increase among the occupations included in that AI and information-technology comparison.
These occupations generally require different skills, education and experience from customer-service, clerical and administrative work. The projections do not mean workers in declining occupations will automatically qualify for technical jobs. They also do not promise higher wages, benefits or available positions in every community.
AI infrastructure may support growth beyond software
BLS expects AI adoption to increase demand for some of the infrastructure needed to develop and operate these systems. The computing infrastructure providers, data processing, web hosting and related services industry is projected to grow 25.1% from 2025 to 2035, adding 120,400 jobs.
Utilities are projected to grow 9.8%, although the sector’s relatively small size means that would amount to about 58,800 jobs. BLS says nearly all of that growth is expected in electric power generation, transmission and distribution, partly because of rising electricity demand that includes power use associated with AI systems and data centers.
What the projections do not mean
The numbers are not a forecast of immediate AI-driven layoffs. BLS says its employment projections describe a potential long-term scenario based on specific assumptions. The agency does not prepare short-term employment projections and does not attempt to predict future business-cycle activity.
BLS also cautions that the precision of the estimates should not be mistaken for certainty. The agency generally applies technology-related adjustments conservatively when there is convincing evidence of a long-term structural change. Its research says technological effects on employment have historically tended to occur gradually, as employers and workers take time to incorporate new tools into business practices.
AI can change the tasks performed within an occupation without eliminating the occupation itself. In some fields, productivity gains could reduce the number of workers needed for certain tasks; in others, lower costs, new products or demand for AI-related services could support additional hiring.
The periods also matter. The main BLS release covers 2025-35, while the detailed AI and information-technology occupation table covers 2024-34. Those periods should not be treated as interchangeable.
How workers, students and parents can use the data
For workers, the clearest near-term message is to watch how automation and AI are changing daily tasks rather than treating the projections as notice of an imminent job loss. Training in technical, analytical, security or software-related skills may be useful for some people, but the appropriate path depends on a person’s education, experience, interests and local opportunities.
The BLS Occupational Outlook Handbook lets readers compare individual occupations by duties, work setting, education, training, wages and projected outlook. BLS says the handbook includes information on about 600 occupations in more than 300 profiles and covers roughly four out of five jobs in the economy.
National projections can help students, parents, jobseekers and workforce planners compare broad trends, but they cannot determine an individual’s job security or predict conditions in a particular state or city. Readers should also compare the projections with current hiring, layoff and unemployment data, which measure what is happening now rather than what BLS expects under a long-term scenario.
What to watch next
BLS updates its projections annually as new data, research and analysis become available. The next important signals will include employer adoption of AI, actual hiring and layoff patterns, changes in the tasks assigned to workers, and whether workforce-training programs create realistic pathways into occupations with stronger projected demand.
Sources
Look for updates to this story
Discover more from Interactive News
Subscribe to get the latest posts sent to your email.