AI Has Not Ended the Offshore Call Center Hiring Boom
New labor data cited by Apollo’s Torsten Slok suggests customer service jobs in the Philippines and India are still expanding despite heavy AI exposure. For managers, the lesson is not that automation failed, but that cheaper, faster service can increase demand for trained human agents.
Employers building AI into customer service are facing a more complicated reality than the simple story of bots replacing people. The roles most exposed to automation, especially offshore call center jobs, have not collapsed. In key outsourcing markets, they have continued to grow.
That matters for managers deciding how to redesign service teams. AI may reduce the cost of handling an individual customer interaction, but lower cost can also encourage companies to serve more customers, open more support channels and expand into markets that were previously uneconomic.
In September 2025, Salesforce CEO Marc Benioff said the company had cut 4,000 customer service positions. “I need less heads,” he said at the time. That example captured the view held by many executives, that agentic AI, meaning software that can take actions toward a goal with less step-by-step instruction, would sharply reduce demand for support staff.
The labor data cited by Apollo chief economist Torsten Slok points in another direction.
Hiring has kept rising where automation looked most likely
Using figures from the IT & Business Process Association of the Philippines, Slok noted in a recent blog post that call center employment in the Philippines rose every year from 2016 through 2025. Over that period, the workforce nearly doubled to 2 million people.
He also pointed to unemployment trends in two major offshore services markets. From 2021 to July 2026, unemployment in the Philippines fell from 9% to about 5%. In India, it declined from around 7% to 6%.
“If AI were displacing white-collar work at scale, you would expect to see it first in the Philippines and India, where business process outsourcing (call centers, IT support, back-office processing) accounts for a large share of employment,” Slok wrote last week. “Instead, the unemployment rate in both countries has continued to trend lower.”
The Philippines became the world’s largest call center employer about 15 years ago, moving ahead of India. The industry’s growth dates back to the late 1990s and early 2000s, when U.S. and other multinational companies accelerated offshore customer support to reduce costs.
The wage gap remains large. Filipino call center employees earn from 15,000 to more than 120,000 Philippine pesos a month, or roughly $243 to $1,948. In the U.S., Indeed puts the average monthly pay for call center representatives at about $2,866.
At the same time, the work is highly exposed to automation. The Brookings Institution has estimated that 86% of customer service representative tasks have high automation potential.
The Jevons paradox explanation
Slok argues that the apparent contradiction is an example of Jevons paradox, an economic idea from 1865. William Stanley Jevons observed that James Watt’s more efficient steam engine made coal more useful and cheaper to apply, but coal consumption rose instead of falling.
“This is Jevons paradox in action,” Slok wrote. “As AI makes call center work cheaper and faster, companies are buying more of it, not less.”
His interpretation is that AI can lower the cost of professional tasks, which may raise demand for the service overall. In customer support, that might mean more conversations, more languages covered, longer service hours or new customer segments.
“Lower cost per interaction does not mean fewer interactions,” Slok said. “It means more customers served, more channels opened and more markets worth reaching. The technology that was supposed to shrink the industry is fueling its expansion.”
Similar dynamics have appeared in other professions. Geoffrey Hinton, often described as a “godfather of AI,” said a decade ago that radiology would eventually not need to be performed by humans because the work could be automated. Instead, Christoph Herpfer, an economist and business administration professor at the University of Virginia’s Darden School of Business, told Fortune that the number of radiologists in the U.S. has risen 10% over the past decade.
“We actually have a huge shortage of radiologists. So the exact opposite of this prediction has happened,” Herpfer said.
What this means for human-led service teams
There is evidence that AI can lift output inside call centers rather than simply remove people from them. A 2023 study led by Erik Brynjolfsson, director of the Stanford Digital Economy Lab, examined more than 5,000 customer support agents and found that an AI conversational assistant increased productivity by an average of 14% per hour.
Brynjolfsson’s earlier research found that an early AI translation feature on eBay increased international exports by 17.5%. Emma Harrington, an economics professor at the University of Virginia, told Fortune the same principle may now be visible in service work. “What we’re now seeing in the data might be a modern version of the same phenomenon,” she said. “We can trade labor more easily across countries when language can be traded more seamlessly.”
For employers, the practical takeaway is to plan for blended operations, not a clean handoff from humans to machines. AI can draft responses, summarize cases, translate text and suggest next actions. Human agents still handle ambiguity, emotion, escalation and customer trust.
Cornell University sociologist and information science professor Benjamin Shestakofsky told Fortune that AI is not yet trained well enough for some complex problems. He also noted that even when AI raises the number of cases agents can manage, people have cognitive limits. In some businesses, human service may remain part of the brand promise itself.
The sharper question for leaders is therefore operational. Which tasks should AI handle, which decisions require trained employees and how should workflows protect quality as volume rises? The companies that answer those questions well may not need fewer people. They may need people working differently.
Reported by Hybrion Insights with reference to Fortune AI.
hybrion
Better together. Built for what's next.
Stay updated
Get insights on AI, teams, and the future of work.