World Bank President Ajay Banga has said that artificial intelligence holds transformative potential for farmers and workers in India and other developing nations — without posing the immediate job-displacement threat that advanced economies face. Speaking at Columbia University during a discussion moderated by CNN's Fareed Zakaria, Banga cited a live AI initiative in Uttar Pradesh as proof that targeted, low-infrastructure technology can already deliver real-world gains for India's most vulnerable agricultural communities.

The Uttar Pradesh Model

Banga pointed to a crop-disease diagnostic tool developed in partnership with Google as a concrete example of what he termed 'small AI.' The application enables illiterate women farmers connected to cooperatives to identify plant diseases and access remedies using nothing more than a basic 3G mobile phone.

'So, we give them a tool that she can speak into a 3G phone, point the camera at the back of a leaf of a diseased plant, and it spits back at you that disease, this insecticide for 25 Rupees from your local cooperative. You press the button. It's yours,' Banga said.

He described the application as a democratisation of expert agricultural knowledge. 'Now, you sort of have democratized the access to amazing knowledge for an illiterate woman farmer. And that to me is, again, A.I. at work,' he added.

Small AI vs Large Language Models

Banga drew a sharp distinction between the generative AI and large language model (LLM) debate dominating developed-world discourse and what is actually deployable in emerging economies. He argued that LLMs require five critical ingredients — computing power, energy, data, skilled personnel, and capital — that remain scarce across most of the developing world.

'There are very few emerging markets with that combination of five things,' he said. By contrast, 'small AI' — predictive tools applied to agriculture, education, healthcare, and skills development — can function within the infrastructure constraints that most low- and middle-income countries currently face.

AI's Job Impact: A Tale of Two Worlds

On the question of employment, Banga offered a strikingly differentiated view. He said that in developed economies, as many as one-third of jobs could potentially be affected by large language models. The exposure falls sharply as countries move down the income ladder.

'In the developed world, while one-third of jobs could potentially be impacted by LLM... the moment you get to a middle income country, that number is in the single digits. By the time you get to a low income country, it's zero,' Banga said.

His argument is that developing economies retain significant headroom to expand productivity and employment before encountering the displacement pressures already being debated in wealthier nations.

The Developed-World Counter-View

Greek Prime Minister Kyriakos Mitsotakis, who appeared alongside Banga at the event, offered a more cautionary perspective. He noted that AI is already encroaching on professions such as law, where research and documentation work traditionally assigned to junior employees can increasingly be automated.

Mitsotakis warned that middle-class workers in advanced economies — already strained by decades of globalisation — face a dual squeeze: AI targeting white-collar jobs and robotics displacing what remains of manufacturing employment. 'That doesn't really leave, you know, a lot of room for optimism regarding the future,' he said.

The contrast underscores a growing divergence in how AI's consequences are likely to be experienced across income groups and geographies — a debate that institutions like the World Bank are now placing at the centre of development strategy.

Broader Implications for India

Banga also flagged similar opportunities in healthcare and other sectors beyond agriculture, suggesting that the 'small AI' framework could serve as a scalable development tool across South Asia and sub-Saharan Africa. As global AI investment concentrates in a handful of wealthy nations, his remarks signal a deliberate push by the World Bank to channel a different kind of AI dividend toward the Global South.

How quickly India's cooperative and agri-tech infrastructure can absorb and scale such tools will be a key test of whether the optimism translates into measurable outcomes for the country's roughly 140 million farming households.