AI agents: The silent workforce shift across India's IT services industry

Behind the automation wave, India's IT sector is reshaping talent, delivery and the future of work

When Tata Consultancy Services (TCS) Chairman N Chandrasekaran said at the company’s recently held 31st annual general meeting that the company will have an equal number of artificial intelligence (AI) agents as employees, it offered the clearest signal yet that India’s technology industry is preparing for a fundamental shift in how it builds its workforce.

 

He also added that “the company, and the industry, are unlikely to hire the same number of people because certain portions of the work that is being done will go to AI agents”.

 

The comment came at a time when enterprises, information technology (IT) services players, and technology firms are experimenting with AI agents to automate several tasks.

 

AI agents are expected to become a new workforce layer, forcing companies to rethink the traditional fresher hiring pyramid that powered India’s IT and consulting industries for three decades. Many executives have been talking about the pyramid reshaping into a diamond, requiring fewer people at the bottom and an AI-native experienced layer in the middle.

 

Many of these companies have already slowed hiring this financial year (2026-27/FY27) as macroeconomic conditions remain volatile and revenue growth continues to remain weak, putting pressure on margins. The outlook for FY27 looks bleak, with three of the top five companies yet to put a concrete number around the hiring of fresh engineering graduates.

 

“The old base was built on labour arbitrage, with junior engineers in lower-cost locations doing execution work. That model is changing. Agents can now handle a significant portion of that execution. So if you’re looking for junior Java developers or .NET engineers to do repetitive coding tasks, you will need fewer of them,” said Arun Ramchandran, chief executive officer (CEO) of digital product engineering platform QBurst.

 

The shift is already evident in the hiring numbers.

 

Though TCS said it will hire 25,000 freshers in FY27, it is one of the lowest campus hiring figures since 2020. Similarly, players such as HCLTech and Wipro have not yet disclosed their campus hiring numbers for FY27.

 

Analysts say there is now greater weight in the middle, with experienced professionals who can govern AI systems, review outputs, and ensure quality. This means hiring growth at the bottom of the pyramid is likely to slow materially across the industry. The firms that once needed thousands of fresh graduates annually may need fewer people to deliver the same amount of work. However, demand will remain strong for AI engineers, data specialists, industry consultants, cybersecurity experts, agent architects, and professionals who can redesign business processes around AI.

 

What will become critical is the role of middle managers. While juniors will not disappear, they will become AI-augmented doers capable of delivering the output of several people. Mid-level managers will increasingly become player-coaches rather than supervisors of large teams, while leadership focuses on faster decision-making enabled by AI.

 

Another company that has highlighted the role of AI agents is EY. The firm announced plans to have 100,000 agents in the company by 2028, operating alongside its 400,000-strong workforce.

 

“Over time, as AI evolves, we’ll have to see where we need to make some slight adjustments. But I’ll tell you, humans are still going to be in the loop. We are not at a place where we can completely rely on AI,” Ajay Anand, vice-chairman, global delivery services, EY, said.

 

A report by Cognizant and Pearson reflects this trend and highlights how AI is transforming India’s entry-level workforce at a faster pace than the global average, while also creating new career pathways and urgent skilling challenges. The study found that 37 per cent of entry-level tasks in India are already performed by AI, compared to a 33 per cent global average, with 18 per cent of human resources leaders reporting that AI now handles half or more of entry-level work.

 

Roles such as routine coding, bug fixing, manual testing, basic application support, production monitoring, and simple documentation in the services industry are declining in demand, while AI application development, agentic AI implementation, Cloud architecture, cybersecurity, data engineering, platform engineering, and AI governance are moving up, according to data from TeamLease Digital.

 

“Across industries, the highest-demand profiles today combine three capabilities: domain expertise, AI fluency (the ability to work with AI tools and agents), and business problem-solving skills. This is why many companies are reducing hiring for entry-level execution roles while increasing hiring for professionals who can supervise AI, interpret outputs, manage exceptions, and drive business outcomes,” said Neeti Sharma, CEO of TeamLease Digital.

 

Experts, however, refuse to sound the alarm for India’s software industry, which has created the largest pool of white-collar professionals over the years. One reason is that token or compute costs have soared for many technology product companies recently without showing a material improvement in efficiency. As services companies bring more agents into the system, their operating costs will also rise in an industry that has traditionally been capital expenditure-light.

 

Phil Fersht, founder and CEO of HfS Research, said the economics of agentic delivery can quickly become challenging if organisations simply throw frontier models at every task. The smarter providers are already moving towards model orchestration strategies where smaller, cheaper models handle routine work and expensive frontier models are reserved for complex reasoning and decision-making.

 

For the IT services industry, compute costs fall into three distinct buckets: costs for internal service delivery (using AI to build and deploy software for clients), costs for internal operations and processes, and investment in proprietary intellectual property and solutions being taken to market. Each has a different cost structure and return on investment calculation.

 

“In many enterprise engagements, the compute cost is being borne by the client. Enterprises want their code built in their own environments, using their own model instances, for security and governance reasons. This shifts the cost burden upstream. Service providers are also actively exploring open-source model alternatives to manage escalating costs. You don’t always need the most powerful closed-source model to get the job done. That’s a cost lever the industry is quietly but seriously pulling,” said Ramachandran.

 

While executives acknowledge that AI agents will automate a growing share of routine work, many argue that the bigger shift will be in skills. Santosh Janardhan, vice-president, infrastructure at Meta, said companies are increasingly looking for resilience and adaptability.

 

“Fundamentally, AI is not going to take away your job. Somebody who knows how to use AI will take your job,” Janardhan said, describing AI as a force multiplier that can improve productivity. 

The middle expands- 

Pyramid to diamond 

AI agents shrink entry-level layers as experienced, AI-ready talent moves to the centre 

Freshers face a reset 

Automation is taking over repetitive tasks, reshaping campus hiring across IT services 

Managers become AI coaches 

Middle leaders will guide agents, not just teams 

Skills take centre stage 

AI fluency, domain knowledge, and problem-solving define the next workforce