Google Develops AI-Powered Digital AgriStack for Farm Mapping in India

Google Developing Digital Agri-Stack Using Satellite Imagery for India: Report

The model assigns a unique ID to each farm, similar to Aadhaar, to create a structured agricultural database. Google is building a digital agri-stack using satellite imagery to enhance farm mapping and financial services. It has begun with a base layer designed to help identify farm boundaries, a senior official announced at Mumbai Tech Week, according to a PTI report. Also Read: Google Announces AI Collaborations for Healthcare, Sustainability, and Agriculture in India "Data and analysis of this base layer of the stack can be used in a slew of applications like bettering the subsidy payments, farm insurance or farm loans," Manish Gupta, a senior director at Google DeepMind, was quoted as saying. "...We have built the first such model that using satellite imagery analysis can now start to identify field boundaries based on usage pattern start to identify what crops are being grown and so on," Gupta reportedly said. The model, inspired by UIDAI's Aadhaar, assigns a unique ID to each farm and is designed to serve as the foundational layer of the "digital agri stack," he reportedly added. Gupta emphasised that startups and agri-tech firms can leverage this data to develop lending, insurance, and subsidy solutions. With 40 percent of India's population engaged in agriculture and farm lending estimated at USD 550 billion, the initiative aims to improve access to formal credit and reduce reliance on informal lenders. These comments come at a time when Indian banks, including SBI, are exploring satellite-based agri-lending models to enhance efficiency. Also Read: Everyone in India Can Ask AI Assistants Questions About Health Issues, Says Meta Official: Report Separately, Gupta reportedly highlighted the scarcity of Indian-language content online, revealing that although 10 percent of the world's 8 billion people speak Hindi, it accounts for just 0.1 percent of web content. To bridge this gap, Google has launched the "Vaani" project, collecting 14,000 hours of audio data across 80 districts in 59 languages as part of its first phase. According to the report, he also said that Indians under appreciate the relevance of artificial intelligence (AI) to expand the boundaries of science, and pointed to the work by DeepMind which led to a Nobel Prize in Chemistry.

Google AgriStack To Use Satellite Data to Map Farms in India

Disclaimer: This content generated by AI & may have errors or hallucinations. Edit before use. Read our Terms of use Google India is developing a digital AgriStack using satellite imagery, starting with a base layer to identify farm boundaries. Manish Gupta, Senior Director at Google DeepMind, said this base layer's data will support applications like subsidy payments, farm insurance, and agricultural loans. ".. we have built the first such model that using satellite imagery analysis can now start to identify field boundaries based on usage patterns, start to identify what crops are being grown and so on," Gupta stated at the Mumbai Tech Week 2025. The model designed by Google is based on UIDAI's Aadhaar system, assigning a unique ID to each farm and serving as the foundation of the "digital agri stack," he added. Gupta stated that it is up to startups and agri-tech companies to develop solutions for loans, crop insurance, and subsidies using the data generated by the platform. These remarks coincide with SBI's aim to improve agricultural lending, as SBI Chairman C.S. Setty recently stated that the bank is considering using satellite data along with structured and unstructured datasets to enhance its agri-lending processes. He noted that SBI plans to apply satellite technology and data analytics to assess crop patterns and gather farmer-related information. Google has also launched the 'Vaani' project which collects speech data through voice recordings from over 700 districts. In the first phase, Gupta gathered 14,000 hours of audio data from 80 districts, covering speakers of 59 languages. Previously, farmers and civil society groups criticised AgriStack for lacking consultations and transparency, warning of potential farmer exclusion and other risks. The letter warned that AgriStack's techno-solutionist approach could introduce new issues, including exploitative lending, farmer exploitation by buyers, diminished farmer autonomy, and a lack of transparency in algorithmic decision-making, among other concerns. According to the statement, the lack of transparency in algorithmic decisions "includes the strong possibility of the database leveraging process using AI techniques that has the danger, either inadvertently or intentionally, of control systems becoming autonomous and even misaligned intelligence directing goals that are not set to our purpose." It further warned that the "lack of an AI policy and a legal security net in India enhances such a danger." It added that corporations might view this as another profit-making opportunity, pushing "so-called solutions" that promote the sale of unsustainable agricultural inputs, increasing farmer debt through fintech, and the "increased threat of (land) dispossession by private corporations." Moreover, there were worries about exclusion and other issues tied to Aadhaar's central role in the system. In a 2019 paper by Niti Aayog on Conclusive Land Titling, it stated: "In India, 'land' continues to be of enormous economic, social and symbolic relevance and hence access to the land and its documented ownership is the core of livelihood for the majority of poor specially the most vulnerable groups like marginal farmers, tribes and other backward categories". The report further highlights that poorly maintained land records and a presumptive titling system contribute to a high volume of land ownership disputes, insecurity of tenure, and reduced investment incentives. Additionally, unclear land titles and inadequate record-keeping hinder entrepreneurship and negatively impact the Ease of Doing Business. The report also notes that resolving a land dispute in India takes an average of 20 years. According to a study by the Centre for Policy Research, land disputes account for 66% of civil cases in the country. Last year, the government claimed to have digitised 95% of rural land records in India. However, the digitisation of land records depends on converting existing records from potentially unreliable paper trails, which may not accurately represent physical boundaries. This can lead to disputes, particularly over mutation histories, further complicating ownership conflicts. The absence of safeguards against data manipulation raises concerns about integrity, privacy, and potential misuse of personal information. Moreover, questions remain about the accuracy of digitized maps, as satellite imagery may overlook natural boundary markers and increase the risk of disputes. Publicly accessible digital records could also expose personal information, leading to risks such as consumer profiling by banks and advertisers. Moreover, concerns remain that "bad data" from historical land records may carry over into the digital system rather than being corrected, potentially embedding past inaccuracies instead of resolving them. There is no mention of a grievance redressal mechanism for individuals who may encounter issues with digitised records, such as inaccuracies or disputes arising from the digitisation process, including information on how citizens can address mistakes or seek help. In a recent parliamentary session, the government stated that AgriStack and the State Registry follow a federated architecture, which is why data ownership remains with individual state governments. Another parliamentary query raised concerns regarding the Digital Agriculture Mission (DAM), and the digital divide and infrastructural gaps in agricultural initiatives. Citing NASSCOM data, it noted that only two percent of Indian farmers use agricultural apps, making digital adoption challenging for small and marginal farmers. The government's response, however, largely sidestepped these concerns. Farmer IDs are also dynamically linked to state land records, granting authorities access to details such as demographics, crops sown, and ownership information. All of this narrows down to whether the Digital Agriculture Mission enables extensive state overreach by centralising farmers' personal and agricultural data through a mandatory Aadhaar-linked digital identity.