How Managers in India Are Dealing with AI Overuse

India is a global leader in adopting artificial intelligence at the workplace, but there’s also an element of AI overuse and potentially AI overload that managers have to watch out for. 

“AI tools are very good if you’re using them to accelerate your thinking, but at times people will outsource their thinking to these tools — that becomes a problem,” said Vipul Taneja, co-founder and chief technology officer at Veersa Technologies, a software-services provider in Noida. 

He said the issue comes up more with younger, less experienced employees. While on the one hand, the leadership encourages all employees to use AI to be more productive, they are still figuring out a balance between efficient use and overuse. 

Managers need to be mindful of the potential for AI overload and related stress, said Sheetal Sandhu, Gurgaon-based Group CHRO at ICRA Ltd., a credit ratings and analytics provider. “It’s for real. People are feeling burnt out,” Sandhu said. 

At her firm, she said they are focusing on AI augmentation rather than AI substitution. “We need to maintain our core skills — our analysis skills, our decision-making skills, [and] empathy. All that can’t be outsourced,” she said. 

Toolkit: Using AI for Employment Purposes

Building the AI Use Culture

About 87% of enterprises in India actively use AI solutions, and one recent study showed that 4 out of 5 employees in India use AI multiple times a week. These are “the highest levels recorded across all surveyed markets,” according to ADP, which conducted the study. 

One of the largest adopters of AI in India are companies in the information technology, outsourcing, and related services industry. Leaders at tech companies and startups are using different techniques to create an AI use equilibrium among their teams. 

“My advice to managers would be to have a critical discussion, not just about the AI use policy but also about the AI use culture,” said Tarunima Prabhakar, co-founder at Tattle Civic Technologies in Delhi, a startup that builds citizen-centric technology tools. 

One of the most prevalent uses of AI is to write computer programs or code. However, the final output depends a lot on the context the AI tool is given.

At some large companies, entry-level software engineers are relying heavily on AI without always understanding the concepts or giving the right inputs. “It’s making life miserable for middle management because they are having to review really badly written code,” Prabhakar said. 

That happened at her organization last year with two young women who had been hired to write code. These women, who were 19- and 21-years-old, were hired from a skilling program for rural youth where they were taught to use AI liberally. 

They did so at Tattle, resulting in a poor-quality project. “They were using ChatGPT for very basic things,” Prabhakar said. 

Earlier this year, when Tattle organized its biannual in-person meeting of its remote team, they discussed AI use, its environmental cost, and the potential of cognitive decline. That was illuminating for the over-users. “All of these things are not apparent to them,” Prabhakar said. 

Based on the discussions, the company has now created an AI use policy, to outline what’s OK and what’s not OK to do. 

Research: The State of AI in HR 2026

Show, Don’t Tell

At Veersa, Taneja said they haven’t created a list of do’s and don’ts around AI use, because they fear it may curb employees’ creativity.

Rather they are trying to show by example how people and AI can work together. 

In coding, Taneja sees human need in both providing the right inputs to the AI tool and in reviewing the final work. “AI is as good as the context it gets,” Taneja said. If it gets incomplete information, or something that isn’t articulated well, it will hallucinate. 

To address this, the Veersa team has built in some tools within their technology systems to catch the obvious mistakes that a new software engineer relying solely on AI might make. In addition, they added steps for human-in-the-loop where a human is required to review the AI’s work even during the project. 

They created a prototype with these practices and presented it as a case study to the leaders of different projects, Taneja noted. He said other project managers came forward to adopt this approach. “Slowly we are increasing the adoption,” Taneja added.

He emphasized the need for team members to verify the final result, and the need to understand why and what AI has produced. “At the end of the day, ownership remains yours,” he said. 

Newsletter: Subscribe to SHRM Workplace Compliance News

‘Evolving Situation’

At large technology companies, where management is pushing aggressively for AI use, team leaders say there are no standard parameters on where not to use AI. 

“What control you hand over and what you retain — that’s a balance we’re all trying to learn,” said Anubhuti Varshney, a Bengaluru-based senior information technology professional at a major U.S. technology corporation.

When team members leave everything to AI, more than they should, she uses a direct approach to discuss the issue. 

She related one instance with an intern who was given a project. 

“Things that we would have thought she would do in a month, she did in a week. But then it was all wrong,” Varshney said. They realized that the intern had used AI to do everything, from design to coding to data analysis. 

Varshney asked the intern to think out her approach in advance, as the end result didn’t make sense. She said there is a scope for AI overload to cause stress, as organizations use AI to push for faster execution. 

Management now has “a new way to ask you to crunch your timelines,” Varshney said, adding that she too has been asked to move up a deadline for a project. 

“This is an evolving situation,” she said.

Shefali Anand is a New Delhi-based journalist and a former correspondent for The Wall Street Journal.