American companies have continued hiring offshore customer support workers at accelerating rates despite widespread predictions that AI would eliminate these roles entirely. Call centre employment in the Philippines nearly doubled between 2016 and 2025 to approximately 2 million workers, whilst Philippine unemployment fell from 9% in 2021 to 4% by March 2026. This expansion directly contradicts the narrative that dominated tech leadership discourse—Marc Benioff's 2025 announcement that Salesforce had cut 4,000 customer service jobs whilst deploying AI agents suggested a clear trajectory toward automation-driven workforce reduction. Yet the data reveals the opposite pattern: as AI makes customer support cheaper and faster to operate, companies are expanding their service capacity rather than contracting it. Apollo chief economist Torsten Slok characterised this as "Jevons paradox in action"—lower cost per interaction does not translate to fewer interactions, but rather more customers served, more channels opened, and more markets worth reaching.
For CX professionals managing teams across Zendesk, Salesforce Service Cloud, and similar platforms, this trend carries immediate strategic implications. The assumption that AI-driven automation would reduce headcount and consolidate support operations has proven incorrect; instead, teams are operating in an environment where human agents and digital agents coexist within expanded service ecosystems. This raises a critical question: are your current staffing models and agent training programmes built for this hybrid reality, or are they still structured around the automation-replacement narrative? The data suggests that competitive advantage now lies not in choosing between human and AI agents, but in orchestrating them effectively across more touchpoints and customer segments than previously economically viable.
The pattern extends beyond customer service into other white-collar professions. Radiology, once predicted to become fully automated, has instead seen a 10% increase in practising radiologists in the United States over the past decade, with persistent shortages across the sector. This consistency across industries suggests that technology expansion, rather than displacement, may be the structural outcome when automation reduces operational friction. For support leaders, this means the strategic question is no longer whether to invest in AI tooling—that decision is effectively made—but how to scale human expertise and contextual judgment alongside those tools to capture the expanded market opportunity that cheaper, faster support creates.
How the industry that AI supposedly 'killed' has not stopped American companies from hiring abroad, and w The Times of India