AI not going to eliminate us; we'll do more & be busier: Cognizant's Babak Hodjat

Bengaluru: Cognizant Chief AI Officer Babak Hodjat, who has been in the field of Artificial Intelligence for over 30 years, has said AI is not going to eliminate us, and that we will do more and be busier. Hodjat spoke to DH on the sidelines of the recent India AI Impact Summit 2026. He was in Egypt the previous week and went to San Francisco and San Diego, before attending the summit in New Delhi. He said the expectation was to be in all these places, because he had “work-related stuff” to do. “It has nothing to do with AI. It’s just that the technology has come to a point where we expect more, and we do more. About 150 years ago, we didn’t have the means to do this. If someone wanted to be at all these four different places, it would have taken them three months just to get around. So the expectation was lower,” he said.

Hodjat, who is the AI pioneer, said he knows it is kind of counterintuitive, but each time a technology like this has come along that made us more productive, it also made us busier. “This is no exception, because this technology is not going to eliminate us. It’s an intelligence that’s very different from human intelligence, and therefore, the two have to work together. The intentionality is still with us; the goal-setting is still with us. Much of the responsibility is still with us, and so our role is there, but we can do more,” he explained.

AI is useful and many people are starting to use it that the use cases are going to proliferate, and we will see AI in places that we didn’t expect. Right now, we are scratching the surface and a lot of the use cases look like chat bots, very standard, or maybe it’s for coding, but people are starting to experiment and use them in many other areas, he said.

“There is the ‘state-of-the-art’ in AI that is indeed changing and being pushed out every few months. Today, you have long running reasoning, agentic systems that are just autonomously, or semi-autonomously running and doing stuff. You set them out, and have them go off and do something, and then they come back and report to you every once in a while. This is new,” he added.

According to Hodjat, we will have agents running at home and doing stuff for us. “Every once in a while, you can dial in, maybe WhatsApp to them and say, Hey, how are things at home? It might actually ping you back every once in a while, and say, Hey, your kid just came home. Everything’s good. Don’t worry. These sorts of things today are available and possible. But there’s the question of the future. I think the rate of innovation in AI has gone up simply on account of more people paying attention to it. AI used to be a small niche. It was like a side event in the world of computing and programming. Today, it is mainstream, and so not only does it have a lot of users, but there are many researchers coming from different backgrounds who are looking to improve the state of the art. And that’s why we’re getting this higher frequency of innovation in AI,” he explained.

Weaknesses exist

Hodjat pointed out that AI has weaknesses. “It’s a different kind of intelligence and it’s not like ours. You learn the world through lived experience. AI doesn’t and that makes it a different kind of intelligence. And today, we don’t know how to set up an AI that would learn through lived experience and have the same kind of abstractions and intelligence as we do. That’s the weakness. AI is very process-heavy, so it needs a lot of processing capacity and memory. We don’t, as humans, run much more efficiently. These are just some areas that might require complete rethinking of AI, like from the basics, we might have to rethink how we built it,” he said.

Since we will be working with many agents, what if one agent hacks another? Hodjat explained that when the internet came out, for the longest time, people refused to do financial transactions on it, because it was insecure, until we found out how to have secure connections. Same analogy holds with AI agents. Today, it’s the early days when a researcher puts something out there, security is not top of the mind for them. But as these technologies get adopted by more serious enterprises, these problems have to be solved, and the good news is, many have already been solved with the internet or with software in the past, he said.

In December last year, Cognizant opened its India AI Lab and Cognizant Moment Studio in Bengaluru to collaborate with clients and partners to help them move from proof-of-concept to scaled adoption with solutions that are quicker and at the same time usable. Hodjat said a lot of the research in the AI Lab is around reliability, trustworthiness, assessing confidence and uncertainty in AI agents, because it’s demanded by clients. “Without that, you can’t have an agentic enterprise,” he said.