What happened
Fortune reports that offshore call-center employment, particularly in the Philippines, continued to grow despite the increasing use of AI tools intended to automate or augment customer service. The article cites data attributed to the IT & Business Process Association of the Philippines and analysis by Apollo chief economist Torsten Slok.
Fortune reports that call-center employment in the Philippines rose every year from 2016 through 2025 and nearly doubled over that period to 2 million workers. The figures are attributed to the IT & Business Process Association of the Philippines and were highlighted in a recent blog post by Torsten Slok, Apollo’s chief economist. Fortune presents the trend as a counterpoint to expectations that AI would quickly eliminate large numbers of customer-service jobs.
The article also reports that unemployment in the Philippines fell from 9% in 2021 to about 5% in July 2026, while India’s unemployment rate fell from around 7% to 6% over the same broad period. Slok argues that if AI were displacing white-collar work at scale, the effects might appear first in countries with large business-process-outsourcing sectors. Fortune reports his conclusion but does not independently verify the underlying labor statistics or establish a direct causal link between AI adoption and national unemployment trends.
Fortune describes the economic mechanism through the idea of Jevons paradox: when technology makes a service cheaper or more efficient, demand for that service can increase enough to expand overall consumption. Applied to customer support, AI could reduce the cost of handling an interaction, allowing companies to serve more customers, open additional channels, or reach markets that were previously too expensive. On this account, a lower cost per interaction can increase the total volume of work even if each individual interaction requires less human effort.
The report says offshore labor remains substantially cheaper than U.S. labor. Fortune gives a monthly wage range of 15,000 to more than 120,000 Philippine pesos for Filipino call-center workers, or about $243 to $1,948, and cites Indeed for an average U.S. monthly call-center wage of about $2,866. It also cites a Brookings Institution estimate that 86% of customer-service-representative tasks have high automation potential. Fortune notes that the article is a version of a story published May 17, 2026, while the current version discusses data through July and a recent economist’s analysis.
Why it matters
The report complicates the assumption that AI-driven efficiency automatically reduces employment. Lower costs may expand demand for customer-service interactions, while human workers may remain necessary for complex cases, quality control, and customer trust. The evidence does not establish that AI has had no displacement effects.
The report matters because it separates technical automability from actual employment outcomes. A task may be suitable for automation without an employer replacing every worker who performs it. Companies may instead use AI to increase the number of customers each agent can serve, reduce handling costs, or make it economical to offer support in more languages and markets. Fortune reports that this is the interpretation offered by Slok and other economists, but the article does not provide a controlled study showing that AI caused the expansion.
Fortune cites a 2023 study led by Stanford Digital Economy Lab director Erik Brynjolfsson that found an AI-based conversational assistant increased productivity by an average of 14% per hour among more than 5,000 customer-support agents. The article also connects this with earlier research on an eBay translation feature, which it says increased international exports by 17.5%. These findings suggest that AI can complement workers and expand cross-border commerce, although they do not predict how newer agentic systems will affect staffing, wages, or working conditions.
The report also emphasizes limits that could preserve demand for human agents. Cornell sociologist and information science professor Benjamin Shestakofsky tells Fortune that current AI systems may struggle with complex customer problems and that human workers have limits on how much additional workload they can absorb. He also argues that some companies may continue to offer human support as a trust or brand feature. These points are relevant to consumers and workers, but they are expert assessments reported by Fortune, not independently tested conclusions in this article.
For workers, the practical question is not only whether total employment rises, but what happens to pay, job quality, monitoring, schedules, training, and the intensity of each shift. An AI-assisted workplace could preserve head count while increasing expected throughput, or it could support expansion in some markets while eliminating particular tasks and entry-level roles. Fortune’s figures show aggregate employment growth, but they do not answer these distributional questions.
What to watch next
Future evidence should distinguish between job creation, job losses, productivity gains, wage changes, and shifts in the kinds of work agents perform. It will also be important to track whether AI increases total customer-service demand, reduces staffing per interaction, or changes the balance between offshore and domestic work.
The next useful evidence would be firm- or industry-level data comparing customer-service staffing, AI deployment, interaction volume, wages, and hours before and after adoption. National unemployment rates are too broad to isolate AI’s effects because they also reflect population growth, trade, investment, migration, education, and wider economic conditions. Fortune’s report does not provide that more granular comparison.
Watch whether companies use AI mainly as an assistant, as a triage layer, or as an autonomous customer-service system. The employment consequences will differ across those models. Systems that draft replies or retrieve information may raise productivity while retaining human agents; systems that resolve routine cases end to end could reduce demand for particular categories of work. The source provides no verified figures on current autonomous-resolution rates or on how many jobs have been eliminated.
Wages and working conditions will be an important test of the optimistic interpretation. If AI expands demand without improving compensation or job quality, rising employment would not necessarily mean workers are benefiting. Conversely, sustained hiring alongside better pay and lower workload would provide stronger evidence that AI is complementing labor. Fortune reports wage levels and productivity findings, but not a current, comparable wage or workload study for AI-enabled offshore call centers.
The report should also be read against its publication history. Fortune identifies the current article as a version of an earlier May 17 story, so readers should not treat every framing point as newly reported. The concrete newer material described in the source is the discussion of employment and unemployment data through July 2026 and Slok’s recent analysis. Those claims remain attributed to Fortune and its cited sources and are not independently confirmed here.