What Jobs Will AI Replace? A Sector-by-Sector Breakdown
AI is already replacing jobs built on repetitive, rules-based tasks: data entry, routine customer service, basic bookkeeping, first-draft writing, entry-level coding. It is not replacing whole professions wholesale. The Bureau of Labor Statistics 2024-2034 projections show job loss concentrated in specific task categories inside industries.
That distinction matters because most "will AI replace my job" anxiety runs on the wrong mental model. Nobody is eliminating "accountant" or "lawyer" as a category. Employers are eliminating the parts of those jobs that were always closer to data processing than judgment.
This page is a directory. Below is the short answer for what is happening across the economy right now, then links to a deeper breakdown for each major sector.
What the projections actually say
The BLS expects the U.S. economy to add 5.2 million jobs from 2024 to 2034, a growth rate of 3.1 percent, slower than the 13.0 percent growth logged the decade before, according to the BLS Employment Projections release. Growth is slowing, with no reversal in the projections.
Inside that overall growth, some occupations are shrinking. Computer programmer employment is projected to decline 6 percent over the decade, with the BLS handbook stating plainly that "computer programming work continues to be automated, with many companies expected to leverage technologies, including artificial intelligence (AI), to automate repetitive programming tasks." That is one of the few occupations where the BLS names AI as a direct cause in its own published text, rather than commentary layered on top of the data.
Retail trade is projected to lose the most jobs of any sector (-1.2 percent), driven by automation, consolidation, and e-commerce, per the BLS Monthly Labor Review. Clerical and administrative roles show the steepest percentage drops: analysis of the BLS data by Visual Capitalist puts cashiers at a projected loss of 313,600 positions by 2034, the largest raw decline of any single occupation, with office assistants next at roughly 177,800 lost roles and clerical typists facing the steepest percentage decline at -36.1 percent.
Layoffs already tied to AI are measurable, if still a minority cause. Challenger, Gray & Christmas tracked 54,836 layoffs in 2025 where employers cited AI as a reason, and 71,825 total since the firm started tracking that category in 2023. For context, DOGE-related federal cuts accounted for far more (293,753) in 2025. AI is a real, rising cause of job cuts. It is not yet the largest one.
Which tasks are exposed right now
Across every sector, the exposed work shares three traits: it is repetitive, it produces a written or numerical output, and a human doesn't need to be physically present to do it. That covers first-draft writing, transcription, basic data entry, routine scheduling, simple customer support tickets, standard contract review, junior-level code, and first-pass financial reconciliation.
The clearest hard evidence of this shift shows up in hiring, not firing. A Stanford Digital Economy Lab study using ADP payroll data found that workers aged 22 to 25 in the most AI-exposed occupations, including software engineering and customer service, saw a 13 percent relative employment decline after generative AI adoption spread, a figure that grew to roughly 16 percent by October 2025 with no reversal yet visible. Workers in the same occupations with more experience did not see the same drop. Employers are not firing veterans. They're not hiring as many juniors to replace the entry-level tasks AI now handles.
Which tasks stay protected, and why
Work that requires legal accountability, licensure, physical presence, or a human to own the consequences of a decision is not going away. A radiologist's read gets a second look from AI tools, but the license, and the liability if something is missed, still sits with the physician. A contract needs a licensed attorney's signature for it to hold up, not just correct language. A plumber has to be in the building. A teacher managing a classroom of 25 kids in real time is doing something no chatbot can do regardless of how good its lesson plans are.
The protected work has a common feature: someone has to be answerable for it in a way that can't be automated away. Insurance, licensing boards, and courts all still require a specific accountable human, not a vendor's model.
What is actually happening in workplaces already
The clearest, most-cited real-world evidence is the entry-level effect documented above by Stanford. It is measured hiring behavior already showing up in payroll data through late 2025.
At the same time, the Yale Budget Lab's ongoing analysis found no broad, economy-wide labor market disruption in the first 33 months after ChatGPT's release, and no clear relationship yet between an occupation's AI exposure and its unemployment rate through August 2025. Both things are true at once: entry-level, AI-exposed roles inside specific occupations are shrinking, while the broader labor market hasn't cracked.
Why growth and decline show up in the same sector
One reason the "will AI take all the jobs" framing misleads people is that growth and decline coexist inside the same industries. IT and computer occupations broadly are still projected to grow through 2034, even while computer programmers specifically decline, according to the BLS Monthly Labor Review. Demand for AI-adjacent roles, people who deploy, monitor, and govern these systems, is rising even as the narrower, most-automatable task inside the same field shrinks. The same pattern shows up in retail: overall retail trade jobs decline, but roles that require in-person problem solving or specialized product knowledge are more insulated than pure checkout and stocking positions.
This is why a sector-level "safe or not safe" verdict is close to useless on its own. Two people in the same company, same title even, can have very different exposure depending on which slice of the job they actually spend their day doing.
What to do about it
Don't try to out-compete AI at the task level; move toward the parts of your job it can't touch. If your role includes both repetitive output and judgment calls, document and advertise the judgment-call work specifically in reviews and resumes, since that's the part that survives restructuring. If you're early-career in a heavily AI-exposed field like software or entry-level analytics, the Stanford data suggests actively seeking roles or responsibilities that put you in front of clients, decisions, or ambiguous problems, not just execution queues. Ask directly in interviews how a team uses AI day to day; teams already integrating it thoughtfully are less likely to do a panic-driven headcount cut later.
Track your own task mix over a normal week. If more than half your time goes to work that fits the exposed pattern (repetitive, written or numeric output, no physical presence required), that's the signal to start shifting responsibilities deliberately, while there is still time to do it on your own terms. Most restructuring happens gradually, through attrition and hiring freezes on the exposed tasks.
Explore your sector
Each sector faces this differently. Go deeper on the one that applies to you:
- Will AI replace healthcare jobs?
- Will AI replace finance jobs?
- Will AI replace tech jobs? (coming soon)
- Will AI replace creative jobs?
- Will AI replace legal jobs?
- Will AI replace engineering and skilled trades?
- Will AI replace business and office jobs?
The tasks you keep decide how replaceable you are
The tasks you can still do without leaning on AI are what make you hard to replace here. The free 5-Day AI Reset is a five-email course built around exactly that: Day 2 has you take one task back and do it unassisted. One small change per day, and it stays useful no matter which way the job market moves.
FAQ
Is AI actually replacing jobs right now, or is this still a prediction?
Both. Measured layoffs tied to AI reached 54,836 in 2025 per Challenger, Gray & Christmas, and entry-level hiring in AI-exposed roles has already dropped per Stanford's payroll-data study. Broader economy-wide disruption, per Yale's tracking, hasn't shown up yet.
Which jobs are safest from AI?
Jobs requiring licensure, physical presence, or personal legal accountability: surgeons, electricians, plumbers, and licensed attorneys signing off on filings. See jobs that AI can't replace for a fuller list.
Which occupation is BLS projecting to shrink the most in raw numbers?
Cashiers, with a projected loss of about 313,600 positions by 2034, according to BLS data analyzed by Visual Capitalist.
Are entry-level jobs disappearing faster than senior ones?
Yes, in AI-exposed fields specifically. Stanford's study found a 13-16 percent relative employment decline for workers 22-25 in AI-exposed roles, with no comparable decline for more experienced workers in the same occupations.
Should I switch industries because of AI?
Not necessarily. The evidence points to task-level exposure inside occupations, not whole-industry collapse. Identify which of your specific tasks are exposed and shift your time toward the judgment-based ones instead of changing fields outright.