This article has been published in partnership with the Gates Foundation.
In Gujarat’s Ajarpura village, Gayatriben knew something was wrong when one of her buffaloes kept banging its head against a tree. In another place called Ishnav (Palaj),Charmiben was trying to understand why a cow was not growing properly even after being dewormed.
For farmers, questions like these often require quick answers. Today, both women can turn to an artificial intelligence system and ask in their native language.
Across India, AI is finding new ways to support farmers, from crop and weather advisories to cattle care and access to agricultural services. The Better Indiaexplored the Gates Foundation’s 2026 Goalkeepers Report, spoke to farmers and technology leaders, and looked at initiatives such as Maharashtra’s MahaVISTAAR and Amul’s Sarlaben to understand how AI is being shaped around farmers’ everyday needs, and what it could mean for making useful, personalised information more accessible at the grassroots.
The right advice can save a whole crop
For Dipali Kalbhor, a smallholder farmer and entrepreneur from Maharashtra, farming has been a journey of setbacks and second chances.
After a failed poultry venture, she rebuilt her livelihood through beekeeping, growing from 30 bee boxes to more than 1,000 today.
Kalbhor tells The Better India that one of her biggest lessons has been that farmers need both a strong team and access to the right information. She discovered MahaVISTAAR during an agricultural training session and now uses it for crop prices, weather, pest control and government schemes.
Its fertiliser calculator has been particularly useful, helping her reduce fertiliser use by around 15%, she says. She also uses the platform to check local crop prices and connect with agricultural officials, calling it a tool she consults “at least two or three times a day.”
She also used it for guidance on raising 10,000 mulberry saplings in a nursery, information about a vermicompost scheme and availing contact with government officers.
How Sarlaben helps dairy farmers
In Gujarat, Amul has introduced Sarlaben, an AI assistant that offers dairy farmers guidance on cattle health, feeding, breeding, vaccinations,offering dairy farmers guidance on cattle health, feeding, breeding, vaccinations and government schemes in Gujarati.
“One of our buffaloes wouldn’t stop banging his head on a tree. Before, I’d have to go through the cooperative, pay Rs 150 for a vet visit, then wait for him to get the issue resolved. Now, I just call Sarlaben,” Gayatriben says.
Sarlaben’s inclusive approach ensures that even farmers with limited digital literacy can benefit from AI-driven insights.
Sarlaben suggested itching around the animal’s horns could be the reason and advised cleaning the area with oil and applying turmeric and oil.
“And it worked!” she says.
Meenaben faced a more serious situation when a cow was about to give birth and the calf was not emerging.
Sarlaben connects with Amul’s databases to provide real-time advice on animal health, breeding, feeding, and veterinary care.
“We thought we’d lose her. She was about to give birth, but the calf wasn’t coming. Panicking, we called Sarlaben.”
The system helped them recognise the seriousness of the situation and connect with a specialist. The calf did not survive, but the cow did.
SarlabenAI helped the farmer understand when specialist intervention was needed.
‘My cows are family to me’
For Charmiben, the challenge is simple: animals cannot say what hurts.
“My cows are family to me,” she says. “But at home, my children can say, ‘Mummy, my stomach hurts.’ My cows can’t tell me when something’s wrong, no?”
So when she is unsure, she says, “Now, I just ask Sarlaben.”
According to Ajay Sheth, HOD IT at Amul, Sarlaben draws on information Amul already holds about farmers and their cattle.
“Sarlaben recognises farmers by their mobile number,” he explains. The system connects their questions to “the data Amul already has about their animals”.
“So it’s never generic advice,” he says. Instead, it can be “specific to your cow, backed by years of Amul’s dairy research.”
Language shapes that experience too.
“When I first spoke to Sarlaben, I thought she was a real person. Her Gujarati was better than mine. I couldn’t believe I was talking to AI!” Gayatriben says.
Meenaben adds, “Earlier, when a cow came into heat, we had to call the cooperative. Now, Sarlaben tracks the heat cycle, shows me technicians, and helps book the experts instantly.”
When one of Charmiben’s cows was not growing properly despite being dewormed, she provided its name and identification number and received guidance on what to feed it.
“And it worked!” she says.
For AI to help farmers, it first has to understand them
The Gates Foundation’s 2026 Goalkeepers Report, Make This Matter: AI, Equity, and the Choice We Can’t Delay, identifies language, local context and access as central challenges to making AI equitable.
More than 90% of the data used to train early large language models came from English-language sources, the report says.
India has 22 scheduled languages, along with hundreds of other languages and dialects.
In agriculture, a misunderstanding can influence what a farmer sprays, feeds, plants or harvests.
The report argues that AI tools need to work in the languages people speak and be built with the people who rely on them.
One of the Goalkeepers Champions for 2026, Prof Ganesh Ramakrishnan, an IIT Bombay professor and a principal investigator and board member of BharatGen, an Indian multilingual AI initiative, says AI cannot simply be built elsewhere and translated later.
Prof Ganesh Ramakrishnan says, it is very important for researchers to innovate with people, and not just for people.
“Does the model think the way an Indian, or more specifically, an Assamese, or someone who is in a remote part ofin South India would think?” he asks.
He describes a woman in a Tamil Nadu kitchen trying to access information while a pressure cooker whistles in the background, showing how technology must work within real lives.
“It’s very important for us to innovate with people, and not just for people,” he says.
He calls the end user a “co-inventor”.
When AI becomes an everyday farming assistant
Debjani Ghosh, a 2026 Goalkeepers Champion recognised for her work on human-centric AI, imagines farmers waking up to weather information, what it could mean for their crop and whether action is needed. Ghosh champions India’s 490 million informal workers—from street vendors to farmers—who are often invisible in global AI conversations and risk being left behind.
“You open your phone, and you have all the information with respect to weather, rain, not rain, what does this mean for your crop, should you take any steps,” she says.
“It will be a voice message; you don’t even have to read, you can listen and understand.”
Debjani Ghosh, a 2026 Goalkeepers Champion envisions AI that works in farmers’ languages and fits into everyday life.
For Ghosh, that is what it means for technology to become “invisible”.
“We have to make life easy for people,” she says. “We have to make it easy for people to earn a livelihood. We have to make it easy for people to live a better life.”
For farmers, that promise will depend on whether AI can understand their questions, recognise their local context and respond in a language they actually speak. As the Gates Foundation’s Goalkeepers Report emphasises, acting now to build AI that is inclusive and accessible can help ensure that this technology becomes a force for greater equity, and that those who stand to benefit most are not left behind.
