AI and Customer Service Jobs: Why Decline Doesn’t Mean No Openings

Customer service representatives topped WTW’s ranking of jobs in demand, according to Capital Pioneer’s November 2 reporting. Agentic AI led a separate ranking of emerging skills in the same research. [1]

I think that pairing deserves attention from anyone considering abandoning customer service after reading AI layoff news. My judgment is that a shrinking occupation can still offer a reasonable next job. The decision about another application deserves a shorter time horizon than the decision about years of further training.

How a shrinking workforce keeps recruiting

BLS projects U.S. customer service employment to decline 5% between 2025 and 2035, a net decrease of 141,800 positions. Yet it projects about 289,500 openings annually, averaged across that decade. It attributes those projected openings to replacing people who change occupations or leave the labor force. [2]

These measures answer different questions. Employment change describes whether the workforce grows or contracts. Occupational openings also account for departures that need replacements, even when some positions are being eliminated. [3]

Imagine a business replacing some departing representatives while leaving other positions unfilled permanently. It could advertise jobs while reducing its workforce. That hypothetical situation requires neither a hiring boom nor a reversal of an automation plan.

I see a useful basis for hope here. Recruitment can continue while the overall direction remains bad. A worker does not have to prove the whole field is recovering before considering a vacancy. But continued recruitment does not cancel the loss of positions.

The case for applying with experience

WTW’s analysis covered job descriptions from more than 1,500 organizations worldwide during January–June 2026. The publication is recent; the underlying evidence concerns the first half of the year. It cannot establish a November hiring rebound. [1]

I would use its customer service ranking as a reason to investigate opportunities. The reporting provides neither raw demand counts nor a detailed explanation of that ranking. [1] An actual vacancy’s requirements would carry much more weight in my decision about applying.

For someone with customer service experience, another role may use abilities already developed. BLS identifies a high school diploma or equivalent and short on-the-job training as the typical entry route. [2] I would investigate suitable openings before assuming a new credential is necessary.

If a position asks for familiar work and offers training on its systems, I would keep it under consideration. Taking another customer service job can be a practical choice while exploring a different direction. The job’s schedule and development opportunities would matter to that plan.

The higher bar for a new qualification

Someone considering lengthy preparation for a new occupation faces a different decision. They need to consider demand when that preparation ends, alongside the work available today. I would require stronger evidence before committing years of study than before applying with skills already acquired.

WTW’s emerging-skills ranking measures increases in mentions compared with the preceding six months. It does not show that customer service vacancies generally require agentic AI expertise. [1]

My advice is to choose training by the requirements of plausible next roles. Identify what you lack and compare the preparation involved. If a course would delay your search, be clear about which opportunities it could make accessible. I would keep applications for suitable current work moving during that investigation.

A qualification should have a clear connection to work you want and can plausibly reach. A course advertised with the latest AI term still needs that connection. Equally, experience alone is a poor reason to ignore a sustained deterioration in the opportunities available to you.

WTW’s global ranking and the U.S. projections cannot establish why individual employers recruit or isolate AI’s contribution to the projected decline. [1] [2] I would be cautious about anyone using these figures to prescribe the same career change to every customer service worker.

What the openings cannot promise

BLS warns that annual openings are ten-year averages, not counts of vacancies available today. Its projections measure demand without estimating the supply of applicants. Moves between employers within the same occupation are excluded, so the figure is not a total of every customer service hire. [3]

That leaves crucial questions unanswered. How many suitable positions exist where you can work? How much competition will you face? A national projection cannot supply those answers. Turning 289,500 into a promise that finding work should be easy would be dishonest.

Nor does an opening guarantee acceptable conditions or advancement. I would judge a role by the work and terms it actually offers. Persistent recruitment could coexist with a weak career path; these sources do not establish that it does, but they offer no assurance against it.

My expectation is that hiring and contraction can coexist for some time. I would become more pessimistic if employers increasingly left departures unfilled or suitable opportunities became harder to access. Forecasts cannot anticipate every shock, and BLS explicitly cautions against treating its averages as precise annual targets. [3]

I do not think workers owe anyone confidence about the future. For someone unsettled by layoffs, the immediate decision can be smaller: whether a real opening makes use of their experience and helps them continue working. That opportunity deserves consideration without requiring a declaration that the danger has passed.