Google DeepMind CEO Demis Hassabis said in an interview with Moneycontrol at the World Economic Forum in Davos on January 20 that India should focus on applying artificial intelligence rather than building new foundational models.
Hassabis said applied AI is the most appropriate focus for India at this stage and advised working closely with existing underlying model providers. His comments come amid ongoing debate within India’s technology and startup ecosystem over whether the country should invest in developing basic AI models or prioritize the application layer.
The DeepMind chief also said India is a strategically important market for Google, noting that a significant portion of the company’s AI research is done in its Bangalore office. He added that India is Google’s largest market by user base and a key region for expanding its artificial intelligence efforts.
Hassabis further said that Indian users have shown strong enthusiasm for AI adoption and the technology is likely to bring significant economic benefits to the country.
He is scheduled to visit India for the first time next month to attend the India AI Impact Summit scheduled from February 15th to 20th. Hassabis told Moneycontrol that he has never visited India before and is looking forward to attending the summit and meeting the local DeepMind team.
This debate reflects broader disagreements within India’s technology leadership. Infosys Chairman Nandan Nilekani has argued that India should establish itself as a global capital for AI use cases instead of participating in the cutthroat race to build language models at scale. In contrast, entrepreneurs like Vivek Raghavan, co-founder of Sarvam AI, warn that without a sovereign AI model, India risks becoming a digital colony.
AI researchers around the world also emphasize that there are geopolitical interests at play. AI pioneer Yoshua Bengio previously told Moneycontrol in January 2025 that countries that invest in their own AI systems can gain a strategic geopolitical advantage, adding that governments should fund and encourage basic model development.
Over the past year, the Indian government has selected around a dozen companies, including Sarvam, Tech Mahindra, Gnani.ai and Gan AI, to develop foundational large-scale language models as part of its national AI strategy.
