India’s AI Gamechanger: How the Next Digital Revolution Is Unfolding Fast
India Eyes AI Revolution with UPI-Style Public Infrastructure
India could soon witness a major digital transformation—this time, not in payments, but in Artificial Intelligence (AI). Just as the Unified Payments Interface (UPI) changed the face of financial transactions, the country now aims to create a similar magic with AI. This bold vision, discussed by top policymakers and tech leaders at The Economic Times Digital Transformation Dialogues in New Delhi, seeks to build a full-stack AI ecosystem that benefits everyone—from startups to government departments to ordinary citizens.
A UPI-Like Model for Artificial Intelligence
Why India Needs an AI Stack
Rajendra Kumar, Secretary (Border Management) at the Ministry of Home Affairs, emphasized the importance of creating an AI stack as a Digital Public Infrastructure (DPI). This AI stack would allow startups and government bodies to develop and deploy AI applications faster—without needing to build their own large-scale compute power or train complex models from scratch.
“Once we have a full AI stack as DPI, we can see AI reaching every Indian,” said Kumar.
Yotta Infrastructure’s co-founder and CEO, Sunil Gupta, echoed this idea and said that India already has the foundational pieces in place—especially affordable GPU infrastructure needed to train and deploy AI models.
“Just like UPI made payments simple and accessible, AI will transform industries like healthcare, agriculture, and education,” Gupta said.
The Rise of Agentic AI in India
What Is Agentic AI?
Agentic AI refers to intelligent systems that can make independent decisions—essentially “AI agents” that can perform tasks without constant human instructions. Think of them as AI-powered assistants or copilots that act autonomously.
Sanket Atal, Managing Director of Salesforce India, said that the AI world has already entered this phase.
“We’ve moved beyond just generative AI. Now it’s about tying everything together with agentic AI,” Atal said, highlighting how their Data Cloud platform supports this new wave.
BharatGPT and the FOMO Trap
Startups are also diving in. Ankush Sabharwal, CEO of Corover AI, shared that around 3,500 AI agents have already been developed using his company’s BharatGPT platform.
However, he warned about jumping on the AI bandwagon too fast. Many users are still reluctant to pay because they don’t see immediate value, and agentic AI can often cost more than employing a human, he pointed out.
“Don’t use AI just because everyone else is. Be sure it solves a real problem,” Sabharwal advised.
Build It or Integrate? India’s AI Dilemma
Should India build AI capabilities from scratch or innovate on top of existing global platforms?
Sachin Bhatia, cofounder of Exotel, argued that India should own the infrastructure, similar to companies like Nvidia.
“Middle-layer applications are nearly free now. The real winners are those who control the infrastructure or the end applications,” Bhatia said.
Regulating AI – The Light-Touch Approach
What Kind of Regulation Works?
While AI regulation is a hot topic globally, most Indian experts at the panel backed a “light-touch” regulatory framework.
Ashish Aggarwal from Nasscom emphasized clear regulatory intent and accountability throughout the AI value chain—from developers to end users.
“Responsible AI is not just ethical—it’s a business necessity,” he said.
Exotel’s Bhatia added that transparency is critical:
“Never pretend that an AI agent is human. And always keep a traceable log of AI decisions.”
Salesforce’s Atal mentioned how his company uses guardrails to prevent AI hallucinations and toxic behavior, while also allowing users to keep their data private.
Will AI Kill Jobs? Not So Fast
Preparing for Disruption
Job displacement remains a real concern. Exotel’s Bhatia didn’t hold back:
“Let’s be honest. There won’t be that many jobs in the future.”
He urged the government to plan for social safety nets when people are made redundant by AI.
Nasscom’s Aggarwal took a more optimistic view—arguing that while some tasks are being automated, India’s booming demand for services provides a cushion in the short to medium term.
“In the long run, education and entrepreneurial mindsets will be key,” he said.
Real-World Use Cases: From Soft Drinks to Borders
Coca-Cola Bottler Uses AI Agents for Operations
Dilip Kant Jha, CIO at SLMG Beverages (India’s largest Coca-Cola bottler), revealed that his company already uses agentic AI for:
-
Logistics planning
-
Predicting safety incidents
-
Managing inventory
-
Avoiding overproduction
“It gives you real-time insight into what to produce and where to send it. It’s a total gamechanger,” Jha said.
But he cautioned that human oversight is essential to catch errors and avoid biases in datasets.
National Security and AI: Where Privacy May Bend
In sensitive areas like border security, AI could be a powerful tool to predict and respond to threats. MHA’s Rajendra Kumar said AI can help manage borders more proactively and efficiently.
However, he argued that privacy concerns must be balanced with national security needs, and special carve-outs might be necessary for certain use cases.
The Road Ahead: Building Trust, Regulation, and Scale
The message from the panel was clear: India is on the verge of an AI revolution, and just like UPI transformed payments, a public AI infrastructure could democratize access to powerful technology.
Getting the balance right—between innovation, regulation, privacy, and trust—will determine how successful this journey is.
As Salesforce’s Atal put it:
“You don’t need to be an AI expert to benefit from it. Systems should be designed to empower everyone.”
