A summary of EY and CII’s “Is India Ready for Agentic AI? The AIdea of India: Outlook 2026”
If there’s one takeaway from EY and the Confederation of Indian Industry’s latest report on artificial intelligence in India, it’s this: Indian businesses have stopped asking whether AI works and started asking how fast they can deploy it.
From pilots to production
The numbers tell a striking story. When EY surveyed over 200 Indian enterprise leaders, it found that 47% now have multiple AI use cases running live in production — a massive leap from last year, when most companies were still stuck in the pilot phase. Only 10% have achieved enterprise-wide deployment, meaning there’s still a long road ahead, but the direction is unmistakable.
Perhaps even more telling: 91% of respondents said speed of deployment was their single biggest factor in buying decisions. Indian businesses aren’t waiting for perfect technology. They’re building around limitations and moving anyway.
The rise of AI agents
The report dedicates significant attention to what it calls “Agentic AI” — autonomous software systems that don’t just respond to questions but actually execute tasks independently, around the clock, at near-zero marginal cost. Think of them as infinitely patient digital workers who never call in sick or demand a raise.
The promise is compelling: AI agents can handle customer queries, process loans, draft content, and manage supply chains while human employees focus on judgment, creativity, and relationships. In a business environment, that’s a powerful competitive edge. The challenges, however, are real — hallucinations, error cascades, and governance gaps mean human oversight remains essential for now.
Rethinking what ROI means
The report pushes back against the idea that AI’s value can be measured in simple cost-cutting terms. EY argues CIOs should evaluate returns across five dimensions: time saved, productivity gains, cost savings, revenue growth, and strategic differentiation.
Real-world examples support this. Axis Bank reported a 30% uplift in product conversions after deploying AI assistants across its employee network. Hindustan Unilever now processes 25 million shelf images monthly using AI. Bharti Airtel has intercepted over 26 billion spam calls. These aren’t lab experiments — they’re scaled operations delivering measurable returns.
Thinking small to win big
Counterintuitively, one of India’s biggest AI opportunities may lie in going smaller. Small Language Models, or SLMs, are compact, efficient, and deployable on edge devices with limited connectivity — perfectly suited for India’s multilingual, mobile-first population. Indian startups are already building SLMs in Tamil, Marathi, Hindi, and other regional languages, bridging a digital divide that English-first global models cannot reach.
Sovereignty and responsibility
The report also frames AI as a matter of national strategy. Through the INR 10,372 crore IndiaAI Mission — including deployment of 40,000 GPUs and 12 startups building indigenous large language models — India is attempting to own its AI future rather than rent it from abroad. Alongside this, the RBI’s FREE-AI framework signals a shift from ethical guidelines on paper to verifiable, auditable AI governance in practice.
The bottom line from this report: India isn’t just adopting AI — it’s betting its economic future on it.