AI isn’t killing outsourcing, it’s killing the FTE-hours model: HBR
MASSACHUSETTS, UNITED STATES — Harvard Business Review’s analysis by AlixPartners partner Abhinav Agrawal argues that AI is not ending outsourcing — it is ending the labor-arbitrage model that has sold it for thirty years.
According to a report from Harvard Business Review, the foundational economics of definable offshore work, FTE headcount metrics, and long-duration rate cards are what generative AI is dismantling; the outsourcing category itself is not going away, it is being repriced.
Labor-arbitrage logic rests on a 30-year-old outsourcing assumption
The foundational economic logic — ‘if work can be defined, standardized, monitored, and moved to a lower-cost labor market, someone else can often do it more cheaply’ — is the explicit target of AI automation across IT services, finance, HR, procurement, customer operations, legal support, claims processing, and analytics, the sectors Agrawal identifies as most severely affected.
Generative AI automates precisely the work that made labor-arbitrage outsourcing profitable: routine, rules-based tasks that are standardizable, monitorable, and scalable across headcount — exactly the characteristics that made them attractive to send offshore in the first place.
IT services and BPO — the two pillars of India’s outsourcing industry — face the deepest exposure because the largest share of contracted work sits in the task categories AI is now executing at cost per output rather than cost per hour.
“Generative AI is rewriting the build-versus-buy calculus behind modern outsourcing,” Agrawal wrote, citing automation of ‘routine, rules-based tasks’ as the mechanism.
Outcome-based pricing replaces the headcount model as AI compresses margins
Companies are transitioning away from per-seat and headcount metrics toward outcome and capability-based pricing models — commercial structures where the unit of value is the result delivered, not the labor hours consumed.
India’s Nifty IT Index hit multi-month lows in February 2026 and TCS dropped approximately 9% in early June 2026 — equity market signals that investors are already pricing headcount-model compression into India’s largest IT outsourcing firms.
The demand profile replacing FTE-hours outsourcing is for data engineering and systems integration expertise — higher-value capabilities required to connect AI automation to enterprise workflows rather than execute the tasks the AI is replacing.
The transition from long-duration rate cards and offshore headcount to outcome and capability-based contracts is the structural divide that separates outsourcing operators who will grow with AI from those who will be compressed by it.
“The old model…built on offshore headcount and long-duration rate cards…is being undermined,” Agrawal wrote.
For BPO and IT outsourcing operators, Agrawal’s HBR analysis reframes the AI disruption debate: the threat is not to outsourcing as a category but to the commercial model that has governed it for thirty years.
Operators who can price on outcomes and compete on capability rather than labor cost are positioned to capture the repriced outsourcing market; those still selling FTE hours are running out of rate card protection.
