Recently, Milind Lakkad, the chief human resources officer (CHRO) of TCS made headlines when he said that generative AI will not replace people, but act as an ‘AI coworker’. Lakkad stressed that while AI is a good thing to happen for the future, it will strictly work in collaboration with the human workforce rather than replacing them.

And while it was nice to see a senior official of one of India’s major IT giants speak up on this looming AI issue, is the picture as rosy as the companies are trying to project? After all, an AI coworker is a ‘worker’ and if ‘it’ is indeed working, there surely is someone who is not. So if there is a chance of AI replacing people at these legacy Indian IT service companies, what are they doing about it?

Gary Bhattacharjee, VP of data strategy and AI at Infosys, told AIM that coding as a job is very clerical in nature, hence generative AI can make coding commoditised. 

And since it can be commoditised, is there a possibility for new companies to have their IT service business models running entirely on generative AI? Analysts at JP Morgan believe that generative AI can be a “deflation driver” in the near term on legacy services as they compete on pricing, necessitate staff retraining and drive loss of competitiveness. In their report, they claim that ChatGPT is likely to deflate legacy services the most and application services the least.

Will they Replace Us?

It’s not just competitive prices, there is a whole question about whether generative AI can entirely replace humans working in these legacy IT service companies. As per a recent survey, a massive 48% of the IT companies in the United States have replaced their staff with ChatGPT. The survey also highlighted that 63% business leaders believe that ChatGPT will lead to workers being laid off in the near future. 

Sanjeev Azad, vice president of technology at GlobalLogic, believes that with the integration of ChatGPT API Indian IT companies might not require as many employees as needed previously.

Even if we ignore generative AI for a moment, companies like TCS and Infosys are ramping up automation at various levels.  Abhijit Mazumder, chief information officer, TCS, told AIM that TCS has its own automatic travel claim process system, something that Bhattacharjee too had mentioned for Infosys. Mazumder also says that if an employee is pursuing a deal, the machine can study and advise whether he is doing well or needs to do better. Or in a case where one has signed a complicated customer contract, a machine can identify who should be monitoring specific contract obligations. 

Why Indian IT is lying

Be it TCS or Infosys, the companies have been optimistic about the upcoming role of generative AI in backend and have said that it will not lead to job losses. But on the flipside, Mazumder also says that TCS believes “the role of the machine is to perform the task, and the role of the human is to train the machine and handle the exceptions”. And if the role of a machine is to perform a (clerical) task, it’s a no-brainer that it will be better than its human counterpart. Most of the work that Indian IT does is service-based — be it e-governance, financial services, telecommunications, HR management etc — which is easy for an AI model to replicate. 

As per a 2015 study by McKinsey, only 4% of the work activities across the US economy required a “median” level of human creativity. That, according to experts, means that 96% of the total jobs can be performed by machines. This is called RPA (robotic process automation). The RPA market alone is expected to reach $11 billion by the year of 2027. Bhattacharjee too, believes that RPA can help in optimization of the processes eliminating the need of human resources entirely. 

However, Indian IT companies believe that the change is not going to be overnight. And while it might benefit the stocks of companies like Buzzfeed to announce using ChatGPT into day-to-day functioning, legacy IT services companies will take a little more time than that. 

As experts point out, using ChatGPT always leaves sensitive data or IP codes vulnerable. As a result, companies like Amazon and JP Morgon have banned the use of ChatGPT in official use. So, until Indian IT receives its own exclusive deal with LLM models like Github Copilot or CodeGen by Hugging Face, which allows it to implement the models without their IP codes going into open source, the companies are likely to remain tight-lipped about a particular model resulting in job losses.