India’s biggest IT firms are hiring far fewer fresh graduates, as AI changes the skills employers seek. Industry experts tell CNA this could create a shortage of experienced workers in the future unless companies and universities do more to develop young talent.
NEW DELHI: Arpita Dayama, 22, a final-year engineering student, is preparing for campus recruitment knowing that strong grades from a premier institute like the Indian Institute of Technology (IIT) Bombay may no longer be enough to stand out.
Dayama - who hopes to work in product management or analytics - says she is now expected to demonstrate practical skills, from data science projects to work involving agentic artificial intelligence.
“You have to show what you have actually worked on, explain it in simple terms and still be able to go into the technical details when they (recruiters) ask questions,” she said.
AI is reshaping how young engineers enter India’s technology industry, one of the country’s biggest private-sector job creators, which employs an estimated 5.95 million people in 2026, experts told CNA.
“The old model (at Indian IT companies) was to hire graduates in bulk and develop them over time,” said Mansee Singhal, country leader at Marsh Workforce & Rewards, which advises Indian and multinational companies on hiring and talent management.
Increasingly, she said, employers want fresh graduates to “arrive with greater technical depth and a better understanding of real business problems from the outset”.
For instance, Tata Consultancy Services (TCS), one of India's biggest IT firms, hired more than 110,000 engineering graduates four years ago, according to its annual report.
But hiring has since fallen sharply. TCS’s intake in financial year 2026 was less than half its level four years ago, while Infosys recruited about 80 per cent fewer fresh graduates in financial year 2025 than three years earlier, according to company reports.
Speaking at the Bloomberg New Economy Forum in New Delhi on Oct 13, Infosys co-founder Nandan Nilekani said AI was increasingly taking over coding tasks traditionally handled by junior engineers.
He said “new skills” would be needed around client needs, business transformation and managing the technology.
Workforce advisers told CNA that companies are adapting by offering in-house training and working more closely with universities.
Academics said universities must also prepare students earlier by developing problem-solving, decision-making and broader technical skills that were traditionally learnt on the job.
RAISING THE BAR
The shift goes beyond AI automating coding tasks.
Sumit Sarawgi, India head at global management consulting firm Oliver Wyman, said multinational companies’ centres in India are also moving beyond routine back-office work to more complex roles involving innovation and specialised industry knowledge.
Sarawgi said an Oliver Wyman survey found that about 45 per cent of managers in India said AI was taking over tasks that traditionally helped junior employees gain experience.
Routine work such as coding and testing was declining, he said, while demand was growing in areas such as AI, cybersecurity and process improvement.
Sarawgi said multinational firms were still hiring fresh graduates, albeit in smaller numbers, and moving them more quickly into roles requiring both technical and business knowledge.
Workforce advisor Singhal said the shift was also changing what employers look for in young engineers. Instead of focusing on whether a graduate could build an application, companies increasingly want people who can understand a customer problem and then work backwards to a technology solution.
She said that could mean using AI for coding while focusing more time on understanding a particular industry, customer needs and practical challenges.
Major Indian IT firms said they are adapting training for fresh graduates to meet the demands of AI-driven roles.
TCS and Infosys said their graduate training programmes now include generative AI, with TCS emphasising hands-on learning and Infosys focusing on problem-solving and design thinking.
But training has limits.
Kamal Karanth, co-founder of staffing firm Xpheno, said training programmes could help junior employees become comfortable using AI, but “true specialisation” in roles such as machine-learning engineers and AI architects requires years of hands-on experience.
THE MISSING MIDDLE
“Where will the future middle managers come from?” asked Oliver Wyman’s Sarawgi, stressing that companies must invest in fresh graduates to develop the next generation of skilled workers.
A 2026 Zinnov survey of multinational companies’ centres in India found that about 60 per cent preferred experienced hires over fresh graduates for roles in AI, cloud computing and cybersecurity, citing the need for expertise and quicker deployment.
“In the short term, companies may hire experienced people, but over time they will need to develop that talent themselves,” said Singhal of Marsh Workforce & Rewards.
Singhal said some large companies are already setting up internal AI, analytics and compliance academies.
For instance, TCS launched an AI Experience Zone in January 2024 to give employees hands-on access to AI tools, after training more than 150,000 employees in the basics of generative AI.
Similarly, Infosys said that its AI Academy provides company-wide training, including practical applications for clients and internal operations.
Sarawgi said companies should use some of the time freed up by AI to train and mentor junior employees, helping them develop the skills and experience needed for more complex roles.
Yet few companies have plans for how employees should use the time saved by AI, he said.
Citing an Oliver Wyman survey, he said only 12 per cent of respondents in India reported that their employers had such plans, while 48 per cent said they received no guidance.
CAN CLASSROOMS KEEP PACE WITH AI?
Universities must also play a bigger role in preparing graduates, industry experts told CNA.
Singhal said companies were working more closely with universities to expose students to real-world business problems and how clients use technology.
For instance, Indian IT firm Wipro has worked with 646 colleges since FY2021 to train teachers in industry-relevant technologies and provide learning materials. About 75,700 students benefited from the programme in FY2025.
Jagat Sesh Challa, an associate professor at Birla Institute of Technology and Science, Pilani, said the change requires more than adding AI tools to the curriculum.
“Capabilities such as system-level thinking, professional judgement and responsible decision-making, which were traditionally developed after joining industry, must now be cultivated during university education itself,” he told CNA.
He said students should work on more real-world projects, with assessments focusing on their ability to evaluate and improve AI-generated solutions and explain their reasoning, rather than simply producing correct answers.
The challenge is extending these opportunities beyond elite institutions, experts said.
Challa called for wider access to mentorship and virtual internships, alongside better faculty training at Tier 2 and Tier 3 colleges.
Karthik Raman of IIT Madras’ Wadhwani School of Data Science and Artificial Intelligence suggested bringing AI specialists into existing departments to update courses, which he said would be “far cheaper than retraining every faculty member from scratch”.
He said universities could also work more closely with industry professionals to teach courses and supervise student projects addressing real-world problems.
However, Sarawgi cautioned that academic change would take time. Until then, he said, companies would continue to carry much of the burden of preparing younger workers for more complex roles.
“Companies, colleges and the individuals themselves are all equal stakeholders,” Sarawgi said. “The idea now is for these three entities … to basically come together to say how do we shape the next 20 years of India’s workforce.”