New Delhi: Global tech giant IBM is sharply focusing on increasing its role in the Indian technology ecosystem, with a targeted three-year plan to offer generative AI-based services and integration especially for public services.
In an interview with Mint, Ana Paula Assis, senior vice-president and chair for Europe, Middle East, Africa and Asia Pacific at IBM, said IBM views India’s enterprise and public-sector use cases as large business opportunities.
“The public sector is one of the biggest opportunities for deployment of AI. There’s no denying that AI is a transformational technology, but enterprises are so far not spending extensively on AI because they are still waiting and watching the technology. They’re assessing the true returns of investments which could be unclear for some, but in public sector services such as in healthcare, education and governance, there is a massive business opportunity,” Assis said.
“To do this, we now have a three-year plan to integrate and progress the role of AI in software. We’re already working on this, and we’ve identified India as a growth market and one of our largest bets and opportunities in Asia,” she added.
Assis was the third key executive that Mint spoke with on the sidelines of the India AI Impact Summit 2026 who singled out India’s public sector services as a major business opportunity.
On Tuesday, Vinod Khosla, founder of Khosla Ventures and an early-stage investor in OpenAI, said one of India’s largest investment opportunities in AI will come in public sector applications.
“I’m not saying this from a private equity investor’s perspective—for the most part, investments will come through philanthropic efforts. But there’s no denying that implementing AI in education, healthcare and India’s agronomy (agriculture economy) will democratize the technology, and create massive opportunities for many,” he said.
Earliest movers
On Wednesday, Demis Hassabis, founder and chief executive of DeepMind—the Google-owned Nobel-winning research arm, said in a keynote that the company will “open access to its frontier AI models” for India’s public sector bodies such as Anusandhan National Research Fund (ANRF) and Atal Innovation Mission (AIM).
IBM was one of the world’s earliest movers in contemporary AI that can understand, think and reason in natural language, creating Watson as early as 2007. The research project evolved into a commercial project in 2014, with the company investing $1 billion back then to develop it further. Watson’s commercial research development was pushed into the field of drug discovery, but couldn’t find success at scale.
Watson’s key drug discovery tool’s sales stopped officially in 2019, a year after DeepMind created AlphaFold—a similar tool.
In 2024, Hassabis’s AlphaFold won the Nobel for its work in the discovery of protein structures.
According to Assis, IBM is consciously staying away from foundational AI models, at least for now. “We’re not looking to get into building the models themselves, but rather, what we see as a big opportunity is the middleware layer, which is what we are developing right now. We have a lot of existing client relations, and we’re looking to leverage that to build our pitch and play in AI right now,” she added.
However, IBM will consciously steer clear of consumer-end applications of modern AI.
“We’re not in the consumer business. Instead, what we’re focusing on is sovereignty of AI and responsible usage, so that enterprises can maximize the accuracy of AI. Instead of just the large language models, we’re focusing on using very focused, sector-specific small AI models—which will allow businesses to train AI to their exact specific need,” she said.
IBM will also steer clear of investing in data centre operations and opportunities.
“We’ll not be into building data centres for AI. Instead, we help companies build the infrastructure layer that goes into the data centres, and we build the software on top of this infrastructure to help companies deploy cloud services—and we’ll continue to do this only going forward,” she added.