Google DeepMind is expanding its India-origin AI models for agriculture—Agricultural Landscape Understanding (ALU) and Agricultural Monitoring & Event Detection (AMED)—to support farming resilience through satellite imagery-based field mapping and activity monitoring.

AnthroKrishi's India-first models are designed to be globally scalable, with outputs adaptable for different countries and sectors to enable more timely, data-driven agricultural decisions.

Alok Talekar of Google DeepMind notes these models are proving effective across credit, crop advisory, and policy decision-making, reinforcing a strategy of targeted AI solutions that scale from local to global impact.

Context: as hunger and productivity pressures rise, with billions facing food insecurity, global production must rise substantially to feed a growing population, underscoring the need for targeted agri interventions.

Telangana and other regional stakeholders highlight the value of shared agricultural data infrastructure to accelerate AI-enabled innovation across government, universities, startups, and industry, moving toward interoperable data layers.

Overall, ALU and AMED are seen as a pivotal shift toward a shared, AI-powered agricultural data infrastructure that supports smart farming, finance access, policy planning, water management, and climate resilience globally.

ALU and AMED enable measuring climate-smart outcomes, such as methane reductions in rice, by mapping fields and providing real-time water management feedback to scale incentive programs on verifiable results.

Public-sector integration includes Telangana’s ADeX and the Krishivaas app delivering hyperlocal crop stress alerts, weather patterns, and pest updates to millions of farmers; Karnataka combines ALU/AMED with local weather and remote sensing to improve water productivity across irrigated land.

Karnataka’s KWRIS uses ALU/AMED with localized weather and remote sensing to enhance river basin planning and irrigation across 2.6 million hectares.

Telangana’s ADEX piloting ALU/AMED-enabled solutions for grid-scale farmer advisories, weather, and pest information to over 5 million farmers.

FAO, supported by Google.org, plans to integrate ALU/AMED into FAO’s CROPGRIDS data repository to enhance global agricultural sustainability monitoring.

The collaboration aims to automate crop detection and map generation in global statistics, backed by $2.5 million from Google.org as part of AI Collaborative: Food Security.