Every few years, a moment forces a nation to rethink its trajectory. DeepSeek, China’s latest AI breakthrough, has sparked one such reckoning in India.
The question is common: “Why can’t India do what China did?” The answer, as Zoho CEO Sridhar Vembu puts it, is simple: “India can, and it is not that hard.”
For decades, India’s tech industry has thrived as an outsourcing hub, but global AI leadership demands more than service revenue.
It requires fundamental research, long-term investment, and a cultural shift toward innovation. Yet, when conversations arise about India’s lag in AI, they often dissolve into blame—on reservations, on governance, on bureaucracy.
But as Vembu points out, Chinese entrepreneurs never had it easy either. And yet, China is competing head-on with Silicon Valley.
On Reddit, an Indian student at a premier institute made a grim observation: “The ship has long sailed.” He argued that China, despite being behind on hardware, has used aggressive open-source research to undermine American dominance. Indian companies, by contrast, remain stuck in low-risk, service-based AI applications.
“We have never caught any bus, train, or even cycle of the tech revolution,” wrote another user. “All this self-patting is for IT service revenue. In real terms, we are nowhere near what the world is achieving.”
The criticism is sharp but not unwarranted. India’s largest corporations largely avoid research-intensive fields. Government funding for non-pharma research remains scarce. The academic system prizes engineering degrees over foundational scientific inquiry.
Even at Davos 2025, HCLTech Chairperson Roshni Nadar Malhotra acknowledged the gap: “Indian companies are focusing on practical AI use cases, but they aren’t yet at the cutting edge of AI research.”
Meanwhile, DeepSeek has already reshaped global markets. The AI race wiped $1 trillion from Nasdaq, with Nvidia losing nearly $600 billion in valuation—its largest drop in history. The AI assistant even overtook ChatGPT on Apple’s App Store in the US and UK.
