Accelerating discovery in India through AI-powered science and education
Google DeepMind ● Covered by 2 sources
Google DeepMind just struck a big AI partnership with India, covering science, schools, farming and power grids. It's DeepMind's biggest national deal yet, and it's landing right as India hosts the world's top AI summit.
Based on reporting by Google DeepMind — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Google DeepMind is planting a flag in India, and it's doing so with more than a press release. Announced on February 18 by Demis Hassabis, Lila Ibrahim and Pushmeet Kohli, the new National Partnership for AI in India follows similar arrangements already running with the US and UK. This one arrives with real substance attached: research tools, classroom pilots, farming APIs and grid forecasting, all bundled together and timed to coincide with India hosting the fourth global AI summit this week.
On the science side, DeepMind is teaming up with the Anusandhan National Research Foundation to hand researchers access to AlphaGenome, the AI Co-scientist system, and Earth AI, alongside hackathons and mentorship for early-career scientists. India is already the fourth-largest user base for AlphaFold worldwide, with more than 180,000 researchers relying on it, so there's an existing appetite to build on. Google.org is backing that ambition with a fresh $30 million Impact Challenge for AI in science, open to researchers and social enterprises both in India and globally, with winners getting engineering support and mentorship through an accelerator program.
Education is the other big pillar, and the numbers here are striking: India now leads the world in daily Gemini usage among students, according to a survey DeepMind ran with Ipsos. At City Montessori School in Lucknow, a study by Fab AI found that in roughly three out of four Gemini conversations, students were pushing for deeper understanding rather than just grabbing an answer. DeepMind is scaling that pattern through Atal Tinkering Labs, which touches over 10,000 schools and 11 million students, and through a partnership with textbook publisher PM Publishers that will turn two million printed books across 250 titles into QR-code-linked AI tutors. A $2 million contribution to IIT Bombay's Indic Language Technologies Research Hub is meant to make sure all of this actually works in India's many languages, not just English.
The agriculture and energy pieces are less flashy but arguably more consequential day to day. Agri AI models are already feeding into Indian farming through startups and groups like the Council on Energy, Environment and Water, while a company called TerraStack is blending satellite, crop and weather data into localized guidance for farmers. On the grid side, DeepMind's WeatherNext models are being integrated with Open Climate Fix to sharpen renewable forecasting, and early tests showed up to an 8% accuracy bump in wind generation forecasts. That matters a lot for a country that became the third-largest solar generator on the planet in 2023 and is chasing 500 gigawatts of renewable capacity by 2030.
My take — AI-written commentary, not fact-checked reporting
This is DeepMind doing what Google does best: turning goodwill gestures into distribution deals, and I don't say that as an insult. Wiring Gemini into two million textbooks and a curriculum used by 11 million students is soft-power infrastructure building, not charity, and it locks India's next generation of researchers and engineers into Google's AI stack years before they ever choose a cloud provider. Smart move, and probably good for India's science output too — but let's not pretend the $30 million challenge fund is anything but a rounding error compared to the value of that lock-in.
Read more about this at: Google DeepMind