AI must serve humanity, humans should be in control, says Microsoft India chief

As a global debate rages over AI safety, Microsoft India and South Asia President Puneet Chandok on Monday said the tech company strongly believes that Artificial Intelligence must remain focused on serving humanity, with humans firmly in control of the technology.

In an interview to PTI, Chandok said there is “real energy” in India around deploying, diffusing and innovating with AI, and that the country is not just consuming technology but also building it, as evidenced in Microsoft’s own operations here. The company has more than 23,000 engineers across 10 cities in India and the largest R&D lab outside the US, underscoring the nation’s growing role in developing AI technology.

Microsoft has topped up its India commitment with a fresh $17.5 billion investment over 2026-29, taking its total pledge to $20.5 billion. The fresh investment, announced in December 2025, built on the $3 billion announced earlier that year, will support cloud and AI infrastructure, skilling and ongoing operations.

On jobs, Chandok said AI is creating new roles such as AI engineers and agent orchestrators, while also changing the nature of existing functions by taking over basic execution, data collection and other routine tasks. He said the shift is pushing people towards higher-value work, with AI and agents handling more of the “grunt work”.

“The work is changing, not ending,” Chandok emphasised. The Microsoft India honcho asserted that AI must be developed to serve humanity, with humans retaining control over the technology at all times.

He outlined the company’s approach to managing risks from rapidly advancing artificial intelligence.

“I won’t comment on what the labs are saying, but at Microsoft, our belief and our principle is simple, which is AI has to be built in the service of humanity, AI has to serve humanity, and humans have to be in control of AI at all times,” Chandok said.

He further said this would require an open and heterogeneous AI ecosystem with multiple models, adequate testing of AI systems, protection of customer intellectual property, and infrastructure that is secure, sovereign and trusted.

Asked about growing calls for a pause in the rapid advancement of AI, Chandok said it was difficult for him to comment on “pausing or not pausing”, and instead stressed the need to build safeguards “by design” and ensure that technologies capable of changing the world are developed responsibly.

“...making sure we build an open, heterogeneous world, making sure we control the IP and ensure it doesn’t leak, make sure the infrastructure is sovereign, secure and runs sustainably, if we build these guardrails, and that’s exactly what we’re doing at Microsoft...and we’re doing this by design, none of it is being bolted on to what we’re building on our platform,” he said.

He added: “Our focus is on diffusion, building this in a secure, sustainable way. If you’re building technologies that can change the world, you have to build it responsibly, and that’s what we’re committed to doing.”

Microsoft on Monday formally launched its fourth India cloud region, in Hyderabad. The India South Central region - as it is called - joins the company’s existing regions in Pune, Chennai and Mumbai. The Hyderabad facility uses effectively zero water for cooling, relying on air-cooled chillers instead of traditional water-based systems, the company said.

Chandok said India is emerging as a major AI market, talent hub and technology builder, with 92 per cent of Indian knowledge workers using AI meaningfully and 40 per cent of enterprises beginning to scale it.

“We talk about frontier models, we should talk about frontier people...there are frontier professionals using AI with a lot of intensity, frequency, getting multiple agents to work with them, reimagining work, and also keeping their human end of the bargain, and then 40 per cent of enterprises in India are starting to scale AI, so there’s a real energy behind India, in really deploying technology, diffusing technology, we’re also innovating,” he said.

Chandok said India is moving from digital public infrastructure that enabled services such as UPI to ‘AI public infrastructure’, aimed at making AI accessible to a billion Indians.

India generates 20 per cent of the world’s data but has only around three per cent of data-centre capacity, creating what he called the “largest infrastructure opportunity of the decade”. He said capacity could rise from about 2 GW to 12-14 GW in the next few years, positioning India as an AI infrastructure hub for India, Asia and the Global South.