5.6 mn tech jobs at stake: Has AI brought a Kodak moment for Indian IT sector and what it must do to survive

India's software services sector may be at its Kodak moment. Anthropic's disclosure that its Claude Cowork automation platform could manage complex coding tasks with the same skill - and in a fraction of the time - as existing professionals triggered a $300 bn sell-off in IT stocks globally.

Indian IT companies have been left vulnerable after a decade of complacency and prioritising profits over innovation. The new stars of ITverse are likely to be solo entrepreneurs who use AI to deliver innovative solutions a hundred times more efficiently at a much lower cost.

This inflection point has been building for years. The dramatic development of ML was first showcased to the world in 2015 after Google's AI programme, AlphaGo, beat the human champ of the ancient Chinese board game Go, which has more permutations than atoms in the universe. During the game, AlphaGo demonstrated it had the ability to creatively outsmart humans and surprise them, something that was not possible by brute computational force alone.


More than a decade on, Indian IT majors are still grappling with the shift assuring investors they will retrain their workforce and find ways to serve global companies from India. The odds are stacked against this plan. India has always worked on labour cost arbitrage.

Up to 65% cheaper, it made business sense for global companies, based in the US or Europe, to readily outsource their application maintenance jobs to Indian firms. For dull, repetitive tasks, cost arbitrage is a big win. But what if a language model could do the same 100x faster with, say, 20% of the manpower needed? It would chip away the incentive to offshore the task. With this development, the narrative recasts AI as a replacement option, from a peripheral assistant.

Then comes the proposed plan of upskilling the workforce. To be fair, India's education system of rote learning has tended to produce steady minds, not breakthrough thinkers. Predictability of the outsourced tasks, along with our English language skills, gave us a seat at the table. If, however, a machine can do these tasks faster and better, working 7 days a week, you have the perfect employee, with a work ethic Narayana Murthy would applaud.

Except that leaves humans with much less to do. Rising nationalism and unemployment in the Western world will likely put pressure on companies to hire locally, anyway. Besides, big tech companies from Google to OpenAI have spent billions in developing language models and data centres - and to monetise this, they will have to get the large companies to buy their models. The likes of us dropping caricatures of ourselves isn't their revenue model.

This multiple whammy is a headache for Indian IT companies and GoI. They must find ways to demonstrate that the 5.6 mn people in white-collar IT jobs can add more value than machines. It is a talk task, given the acceleration in ML. Of course, a modest percentage of these coders will be humans in the loop who will work with 'AI agents'. But there is little doubt that we would need a new business model for Indian IT firms and their employees to thrive.

AI will eventually democratise coding, just like YouTube democratised broadcasting or Google did with information. Look at what it has done to legacy media. AI will become the new capital. This will open unprecedented opportunities for enterprising Indians to become solopreneurs who use AI tools to solve local problems.

Whether we survive this shift and thrive will depend on how well we overhaul our education system and skills programmes by prioritising reasoning, creativity and project-based learning. Our best-case scenario is to develop an army of AI-driven solution providers.

India's creator economy has shown our youngsters know how to discover audiences and build personal brands. With better foundational learning and AI-driven execution, they could go on to conquer the ITverse.