While the Green Revolution made Indian farms more productive, the next revolution must make them predictive. With volatile weather, shrinking plots, and fickle markets, artificial intelligence (AI) is the new compass that can help Indian agriculture navigate uncertainty with insight rather than instinct.
Readiness begins with first principles. Indian agriculture is dominated by very small and fragmented landholdings, with the average operational holding at about 1.08 hectares as per the most recent Agriculture Census. This strengthens the case for hyperlocal intelligence that adapts to soil, microclimate, and market conditions at the plot level.
The country has begun laying the digital foundation for modern farm intelligence. AgriStack is a farmer-centric digital public infrastructure being assembled around three core registries for farmer identity, geo-referenced village maps, and crop sown information. Together, this provides a verified picture of who farms what and where, supported by consent-based data sharing.
The Union
Budget of 2026 took this foundation a step further with the launch of Bharat VISTAAR, adding the intelligence layer to Indian agriculture. It combines AgriStack records with agricultural practices recommended by the Indian Council of Agricultural Research and delivers advice through simple voice calls and basic phones. In doing so, it treats AI as a public service that any farmer can access rather than a premium feature for high-end users.
There is credible evidence that digital advice can reduce risk at scale. In Odisha, the government’s voice-based advisory service, known as Krushi Samruddhi, helped farmers adopt better practices and withstand weather losses more effectively. The estimated benefit-cost ratios were between $12-$19 for every dollar invested, while losses from pest diseases and extreme weather were reduced by nearly 25%.
Some states are also testing new forms of participatory and AI-driven knowledge systems, such as Tamil Nadu, which partnered with Apurva.ai to create a platform that captures farmer knowledge through web and WhatsApp. Assam, Bihar, Jharkhand, Madhya Pradesh, Maharashtra, Rajasthan and Uttar Pradesh have launched AI-based crop yield prediction pilots to support real-time advisory for farmers.
Meanwhile, states are also pairing advice with deeper spatial intelligence. Madhya Pradesh’s UNNATI initiative combines satellite imagery, drone data, geographic information systems, and positioning tools to map crops and estimate yields with greater accuracy. This strengthens planning and makes insurance and relief workflows more transparent.
The promise of scaling these solutions across India is undeniable, but so are the risks associated with it. Data heterogeneity is a defining feature of Indian agriculture, as the sector does not speak in standard codes. Pearl millet is bajra in Hindi, bajri in Gujarati, kambu in Tamil, and sajje in Kannada. AI systems may misinterpret queries or fail to generalise across regions if such linguistic and regional nuances are not considered. AgriStack’s open schemas and crop registries are an import ..
Another important risk is inequity and exclusion. If advisory tools are designed only for smartphones, they will widen sectoral divides marked by informality and low capital. Bharat VISTAAR can mitigate that risk through voice access and integration with national schemes, provided the platform offers toll-free access and supports local extension instead of sidelining it. The Avaaj Otalo experiment in Gujarat showed how a basic phone advisory can improve outcomes where literacy and connectivity a ..
There is also the danger of technological lock-in and vendor dependence. If a handful of companies, such as Bayer, John Deere, Farmonaut, and DeHaat, control the systems that decide what advice farmers see or what inputs they should use, real decisions will start shifting away from farmers themselves. The antidote is an open architecture with public ontologies and standard application interfaces so that private and public innovators compete on quality while speaking the same common language of c ..
AI governance is another overlooked concern. As AI models need to be updated throughout the season, farmers and administrators need clear attribution, transparent change logs, and easy grievance pathways. They are the building blocks of accountable AI in a sector where decisions affect livelihoods every day.
Lately, India’s spending on agricultural R&D has hovered around 0.03% of GDP, indicating that financing is a foundational constraint. Without a clear step up, the country will not be able to build reliable AI models in the diverse Indian agriculture.
The strategic shift that can reconcile both the opportunities and the risks is to pivot from pilot to policy, towards what might be called public to policy. Models must be trained and tested on real-world datasets already flowing through public systems and feed their insights into one common advisory system. This includes AgriStack’s registries for farmer identity, land, and crop, PM-KISAN for identity verification and direct transfer, PMFBY for enrolment, technology-supported yield estimation, ..
There is a final reason to treat AI as a public good in agriculture. The deepest resilience comes from combining science with lived practice. Product councils for Bharat VISTAAR and allied platforms should include environmental scientists, gender specialists, and farmers who represent different landholding and tenancy categories, so that the problems chosen are the right ones and the advice remains practical and trusted.
If the first revolution multiplied yields by putting science into seed and water, the next one will multiply resilience by putting learning into every decision. Either AI speaks every farmer’s language, learns from every plot, and is accountable to every loss, or it solidifies as a premium tool for a few. The bus is leaving, and India cannot afford to miss it; rather, it should be the one driving it.
Source Name : Economic Times