Prime Minister Narendra Modi during the inaugural ceremony of AI Impact Expo 2026 in New Delhi on Monday. photo Credit:
While appreciating the role of AI in agriculture, especially in improving crop yields and farmers’ income, experts have expressed concerns over the quality of advice as the sector is not regulated and suggested that standards should be developed.
addressing a session on Standards and policy pathways to scale AI-enabled, farmer-centric agri-food systems In New Delhi, Shantanu Chaudhary, former director of IIT-Jodhpur, said: “If AI is driving public decision-making with respect to agriculture and agricultural advice, who will be accountable with respect to the quality of the advice? This is fundamentally important because it impacts the livelihoods of many farmers.”
Along with the AI programme, Union Agriculture Minister Shivraj Singh Chouhan is scheduled to launch the newly announced India-expansion scheme on Tuesday.
Registration Process
Chaudhary said there are a large number of start-ups in the US, where government initiatives are focused on enabling intervention and driving scale through AI. But in India, he said something which is very important to focus on today and that is to bring AI in agriculture sector with proper accountability mechanism.
Elaborating further, he said, “It must be clearly monitored, detected and challenged”. This controversial advice is important because it ensures that the guidance provided is used only for the intended purpose and is not influenced by commercial interests or other unethical interferences, he said.
He said, since AI now comes in models, there should be a process to register the models. “There should be a process and protocol defined in the registration. In the absence of any protocol, if someone says its AI ensures 99 per cent accuracy, it makes no sense as it is in relation to the data collected and how generalizable the performance is across multiple stakeholders, multiple scenarios,” he said.
flow problem
Since agricultural land in India is fragmented, soil properties vary depending on the history of land use. As a result, AI models that make predictions based on land data may experience drift when applied from one plot to another, Choudhary said.
Highlighting that the International Telecommunication Union has very strong standardization on digital agriculture and standards, ITU’s Atsuko Okuda said it has more than 200 AI-related standards already approved and another 200 in the pipeline.
He also said that these standards are important for democratizing AI in digital agriculture as not all countries are equipped to start from scratch. “So, when you have standards, it is easier for countries to pick up where the standards end,” he said, adding that ITU also has standards for very specific areas of agriculture such as smart livestock farming based on IoT.
Raghu Chaliganti of Fraunhofer HHI said India has a large number of small farmers at around 130 million and reaching out to them is a big challenge. “How to bring innovations and this AI technology to their doorstep? Farmers want input costs to be reduced and production to increase and better prices for produce,” he said, adding that AI is promoted as a solution that can reduce input costs and maximize productivity.
Published on February 16, 2026
