India’s nearly $100-billion global capability centre (GCC) industry is at an inflection point where hiring is rising, but “portfolio hollowing” has begun. Artificial intelligence is compressing repetitive functions across customer support, finance, Human Resources, analytics and software operations, even as companies continue to expand and add headcount, experts told ET.Industry experts said AI is shrinking junior-level workflows, reducing standalone procedural functions and pushing multinational companies to redesign the kind of work handled by their India centres. Early industry estimates suggest nearly 18% of India’s GCC work still sits in commoditised, process-heavy functions that are increasingly being automated, embedded into AI-driven workflows or rebuilt by AI-native firms, marking a structural shift in the country’s largest white-collar employment engine.According to a recent report by global consulting firm Zinnov, nearly 55% of global GCC portfolios are exposed to AI-led disruption, with procedural and commoditised work carrying the highest risk. “Our analysis of 1.7 million-plus job descriptions across 60-plus countries puts India’s commodity share at 17.7%, far below competing GCC locations like the Philippines at 40.1% and Costa Rica at 41.1%,” said Pari Natarajan, chief executive officer at Zinnov.The hollowing is most visible in finance and HR, where AI has moved from pilot to production. Invoice processing cycle times have fallen from seven days to one day, financial closing process from 12 days to five days, and headcount requirements for these functions by 75%. HR self-service is now resolving 70-80% of queries without human intervention. In software development, output per developer has improved 40-80%. “These aren’t pilot numbers, they’re production benchmarks,” said Vikram Ahuja, co-founder, ANSR, a GCC enabler.The shift is emerging as a critical transition point for India’s GCC ecosystem, which has expanded to more than 2,100 centres employing over 2.3 million people. Industry experts said the disruption is showing up less through large-scale layoffs and more through changes in the composition of work itself. Over 70% of GCCs are now explicitly tasked with driving their parent enterprise’s global AI mandate, and AI hiring across GCCs has surged 131% year-on-year.Companies setting up new GCCs today are planning for 20-40% fewer employees than they would have three years ago, while expecting more output. Enterprises are increasingly prioritising hiring for AI, data, cybersecurity and product engineering roles, while functions such as marketing operations, customer support, finance processes, procurement, talent sourcing and repetitive analytics face the most exposure to AI-led restructuring.“The signal isn’t shrinkage, it’s recomposition,” said Arindam Sen, partner, EY India. He said companies are moving away from broad-based volume hiring toward specialised, capability-led roles as AI automates routine tasks across finance operations, HR processes and IT support functions. Sen added that operations-heavy work faces direct substitution risk from AI, while engineering-heavy work is seeing augmentation, where productivity per engineer rises and the junior end of the workforce pyramid gets compressed rather than eliminated. “The enterprise priority has clearly moved from how cheaply can this centre run to what can this centre build,” he said.Yet the transformation remains uneven. Only 37% of GCCs are running active AI pilots, and 40% of agentic AI projects are projected to be cancelled due to unclear returns and inadequate governance. Gaurav Vasu, founder of UnearthInsight, a market research platform, said there is still limited evidence globally of wholesale elimination of enterprise functions. “AI is changing how work is delivered and improving output efficiency, while selectively reducing highly repetitive layers of work,” he said. Vasu added that many enterprises are still evaluating the economics and scalability of large-scale AI deployment, and most core functions across compliance, governance, scientific research and product development continue to require strong human oversight even as procedural workflows gradually shrink.India’s structural strengths, including engineering and AI talent at scale and the institutional depth to run enterprise-grade operations, give it a long-term edge. But Ahuja noted that advantage is not guaranteed. “The advantage is structural for those moving up the value hierarchy. It’s temporary insulation for those that aren’t,” he said. GCCs still predominantly running standardised processing, even complex processing, face a different trajectory, he added.