Google DeepMind and ICRISAT Launch AI Genetic Atlas for Sorghum

On October 15, 2026, Google DeepMind, in collaboration with the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), the HudsonAlpha Institute for Biotechnology, and the University of Washington, announced a significant initiative to develop an open-access AI-supported genetic atlas for sorghum. This project aims to enhance global food security by utilizing advanced genetic analysis and machine learning techniques to expedite the creation of more resilient and productive sorghum varieties. The initiative is particularly crucial as climate change poses challenges to food systems worldwide, making the development of climate-resilient crops essential for farmers, especially in vulnerable dryland areas.

Sorghum, also known as jowar, is recognized for its potential as a nutrient-rich crop that can adapt to adverse climatic conditions. The genetic atlas will draw from a representative sample of ICRISAT’s genebank, which houses over 44,000 distinct seeds collected from more than 90 countries, including India. By pairing advanced genetic analysis with AI, the project aims to identify the most promising plant combinations, significantly reducing the time required for traditional breeding methods, which can span decades. This approach is expected to facilitate quicker development of sorghum varieties that can withstand climate variability and improve food security.

The collaboration also includes the establishment of a three-year postdoctoral fellowship at ICRISAT, marking Google DeepMind's first fellowship in India and outside the UK. This fellowship will focus on developing tools to characterize genomes for other important crops, such as pulses, which are vital for agricultural resilience. This initiative aligns with India’s national emphasis on millets, reinforcing local research as a blueprint for global agricultural sustainability.

In addition to the genetic atlas, Google has been actively involved in enhancing agricultural resilience through various technology-driven initiatives. The AnthroKrishi team has developed models that support farmers across 11 countries, providing insights that inform policy decisions and innovative practices. Initiatives like the integration of agricultural model APIs into government platforms have enabled widespread dissemination of critical information to farmers, helping them optimize crop production and manage resources effectively.

Looking ahead, the partnership between Google DeepMind and ICRISAT represents a significant step toward leveraging AI for agricultural innovation on a global scale. By creating accessible resources like the genetic atlas, the collaboration aims to empower researchers and breeders worldwide to enhance crop resilience. This initiative not only addresses immediate agricultural challenges but also sets a precedent for future partnerships that harness technology to promote sustainable farming practices and food security in the face of climate change.