In January 2026, Infosys announced it would not hire freshers from its 2024 campus recruitment drive until Q3 FY2027 — a delay of over two years affecting tens of thousands of engineering graduates who had received offer letters and were waiting to join. The official explanation cited project pipeline delays and utilisation optimisation. The unofficial context — widely discussed in India's technology industry — is that AI-driven automation is reducing the entry-level coding, testing, and data processing roles that historically absorbed large cohorts of fresh graduates, and that Infosys, like its peers, is navigating a structural transition in its workforce composition that the traditional mass hiring model cannot accommodate.

The Infosys hiring pause is a single data point in a structural transformation that is reshaping India's largest formal employment sector. India's IT-BPM industry employs approximately 5.4 million people directly and supports tens of millions more in indirect employment. It is the sector that created India's modern middle class, that built the economic foundation of Bengaluru, Hyderabad, and Pune, and that has been the primary engine of India's services export growth for three decades. What happens to it in the AI age is not a sectoral question. It is a national economic question.

The Disruption Dimension

The specific threat that generative AI poses to India's IT sector is concentrated at the entry and mid-level of the skill distribution. The tasks that junior developers, BPO agents, data entry operators, and basic software testers perform — the tasks that have historically been the volume base of India's IT workforce — are precisely the tasks that large language models and AI coding assistants are most capable of automating.

GitHub Copilot, Amazon CodeWhisperer, and Google's Gemini Code Assist are already deployed across India's major IT firms, with productivity assessments showing 30–50% reductions in time-to-completion for standard coding tasks. That productivity improvement has a workforce implication: a team of 10 developers with AI assistance can produce the output that previously required 15, which means that as existing contracts are renewed and new projects are scoped, the headcount required to deliver the same revenue is structurally lower.

The BPO sector — which employs approximately 1.4 million people in customer service, data processing, and business process functions — faces a more immediate disruption. AI-powered customer service agents, automated document processing, and intelligent workflow orchestration are reducing the labour input per unit of BPO output faster than the sector's natural revenue growth can absorb. Several large BPO operators have publicly guided for flat or declining headcount even as revenues grow — a divergence between revenue and employment that has not been seen in the sector's history.

The Opportunity Dimension

The disruption narrative, while real, is incomplete. AI is not simply destroying IT jobs. It is restructuring the skills and roles that the sector demands — creating significant demand for AI engineers, prompt engineers, AI safety specialists, data scientists, MLOps practitioners, and the kind of human-in-the-loop roles that AI systems require for quality assurance, ethics review, and complex decision support.

India's competitive position in this restructured landscape is genuinely strong. Its engineering talent base — 1.5 million STEM graduates annually, with the world's largest pool of English-speaking technical professionals — is the raw material for AI-era IT leadership. The companies that are best positioned to capture the AI transition in IT are those that can move their workforce from execution-focused roles to design, architecture, and AI integration roles — and India's talent pipeline is well-suited to that transition if the training and upskilling infrastructure can be mobilised at scale.

The domestic AI opportunity is equally significant. India's AI startups — Sarvam AI, Krutrim, and a growing cohort of sector-specific AI applications — are building products for India's specific conditions: 22 official languages, low-bandwidth connectivity, and a public sector that manages services at a scale no other country matches. Every government service — from income tax filing to health insurance claims to agricultural advisory — is a potential AI application with a domestic market of hundreds of millions of users and an export market across the developing world.

The Policy Response India Needs

Three policy interventions are immediately necessary.

The first is a national AI reskilling programme at scale — not the existing Skill India framework retrofitted with an AI label, but a purpose-built programme that partners with India's IT majors, its IITs and IIMs, and global AI companies to deliver industry-validated AI skills to 10 million workers over five years. The cost is significant. The cost of not doing it is higher.

The second is social protection for IT sector transition — portable health insurance, reskilling allowances, and income support for workers displaced by AI automation. The IT sector's workforce has historically operated without the social safety nets of manufacturing — no gratuity for contractual workers, no union representation, no sector-specific welfare fund. The transition creates a case for a purpose-built IT sector social security framework.

The third is education system reform — specifically, updating engineering and computer science curricula across India's 4,000+ engineering colleges to include AI fundamentals, machine learning, and the human-AI collaboration skills that the industry now requires. A graduate entering the workforce in 2028 without these skills is entering a labour market that has already moved past them.

AI will disrupt India's IT sector. It will also create the opportunity for India to move from the world's IT back office to the world's AI innovation hub. The difference between those two outcomes is policy.

The Hind covers policy, power, and strategic affairs from India's perspective. Views expressed are analytical and editorial.