IBM’s Patel Warns India Must Train 150 Million More Workers in AI or Risk Losing Its Edge

India counts roughly 600 million workers. Only about 200 million possess enough AI knowledge to matter in the economy taking shape. Sandip Patel sees both the opening and the risk. As managing director of IBM India and South Asia, he told Reuters that turning those 200 million into 350 million represents a “phenomenal opportunity.” The target demands action before the decade ends. Nothing less will position the country as the world’s AI skill capital.

Patel’s arithmetic comes straight from a joint study by IBM’s Institute for Business Value and IndiaAI. Released this month, the analysis projects AI could add more than $500 billion to India’s economy by 2030. But first the share of AI-literate technology workers must climb from 30 percent to nearly 57 percent. That gap spans 150 million people. And the clock shows less than five years remain.

The demographic window narrows fast.

More than half of India’s 1.4 billion citizens sit under age 30. The nation graduates millions of engineers annually. Many feed the global IT services machine that made India the world’s back office. Those same roles now face direct pressure. Generative AI handles coding, ticket resolution, basic analysis. Tasks once measured in headcount now run on model calls. Productivity rises. Job descriptions shift. New competencies appear overnight.

Patel described the dynamic without alarm. “AI is both creating productivity improvements, which is changing the complexion of jobs, but it’s also creating new skill sets that people have to adapt and learn, which then creates newer jobs,” he told ANI at the report launch. The report itself takes a sharper tone. Seventy-two percent of surveyed organizations admit they trail global peers on AI adoption. Only 15 percent scale the technology through cross-functional investment. The rest remain stuck in pilot projects.

India does not stand alone in this pattern. European statistics show just one-fifth of EU enterprises actively use AI, with skills shortages ranking high among barriers. Yet India’s situation carries unique weight. Its young population offers scale few nations match. The challenge lies in converting that headcount into capable talent before automation erodes existing advantages.

Recent data reinforces the tension. A Bain & Company analysis, cited by Times of India in March 2025, forecasts AI-related job openings in India could exceed 2.3 million by 2027 while the talent pool reaches only about 1.2 million. The resulting shortfall surpasses one million professionals. ManpowerGroup’s Global Talent Shortage Survey from February 2026 found 82 percent of Indian employers struggle to fill roles, well above the global 72 percent average. AI skills top the list of unmet needs.

Industry bodies paint a similar picture. Nasscom projections indicate 60 to 65 percent of the existing workforce will require reskilling or upskilling within five years to stay relevant. Current readiness hovers between 20 and 25 percent in many assessments. The gap appears widest outside major tech centers. Government programs such as IndiaAI FutureSkills aim to push data and AI labs into Tier 2 and Tier 3 cities. Progress remains uneven.

IBM has stepped forward with one of the largest corporate commitments. In December 2025 the company pledged to skill 5 million Indian youth by 2030 in AI, cybersecurity and quantum computing. Delivery runs through the IBM SkillsBuild platform, which offers free courses, certifications and project-based learning. Partnerships with the All India Council for Technical Education expand reach across education levels. Patel’s team also grows physical presence beyond traditional hubs. IBM’s Kochi workforce reached nearly 4,000 employees within two years. A new outpost opened in Lucknow. These moves target fresh talent pools and support clients in less saturated markets.

Yet corporate training alone cannot close a 150-million-person deficit. Patel called for coordinated effort across government, companies and academia on both skilling and policy. Stronger intellectual property protections rank high on his list. Without credible safeguards, India risks training engineers who build value for others while domestic firms capture limited returns. “India will need stronger IP enforcement if it wants to move from running the world’s back office to creating monetisable technology of its own,” he noted. Models may originate elsewhere. The workforce that fine-tunes, deploys and innovates on them could sit here. Ownership of the resulting economic upside remains an open question.

Executives inside Indian firms echo the concern. Many report AI pilots that never scale. Budgets favor experimentation over integration. Talent poaching intensifies for the small cadre of senior AI engineers capable of building core systems. Estimates place that elite group below 2,000 people even as overall AI-linked job postings surged 21 percent annually since 2019. Salaries for those roles rose 11 percent each year. Demand outruns supply at every layer.

Policy makers have responded with declarations and frameworks. The AICTE labeled 2025 the “Year of AI” and instructed engineering colleges to embed the subject in core curricula. The IndiaAI Mission channels resources toward future skills. Still, implementation lags. Enterprises surveyed by various reports cite unclear pathways, inconsistent quality and limited focus on applied, project-based learning. Abstract theory does not translate easily into workplace productivity.

Global examples offer partial blueprints. IBM itself automated portions of its human resources function, replacing roughly 200 positions with AI systems that now manage over 1.5 million employee conversations annually. The company redirected displaced workers into new strategic roles rather than simply cutting headcount. Similar transitions appear in manufacturing and services worldwide. Success hinges on deliberate retraining, role redesign and cultural acceptance of continuous learning. India must accelerate that playbook at national scale.

Patel’s message carries measured urgency. The demographic dividend exists today. It will not last forever. Failure to act leaves India supplying cheap, augmented labor while others own the models, the platforms and the margins. Coordinated skilling at speed, paired with sound policy on IP and data, could flip the script. The numbers do not lie. Two hundred million becomes 350 million only through sustained, collective focus. The window stands open. How wide it remains depends on choices made in the next few quarters.