The Indian IT sector is not undergoing a simple return of mass layoffs. The more consequential shift is quieter: companies are reducing routine and coordination-heavy work, leaving some vacancies unfilled and hiring selectively for artificial intelligence, cloud, data, cybersecurity and automation capabilities. The result is a labour market where employment can still grow while many established roles become less secure.
That distinction matters for India’s urban economy. Technology employment is concentrated in major cities and has supported office districts, rental housing, transport networks, education choices and household consumption. A change in the composition of technology work therefore reaches beyond corporate staffing. It affects how workers build careers, how families plan incomes and how cities that depend on high-value services absorb new labour.
The evidence cited in the report points to a transition from headcount-led growth to productivity-led growth. Neeti Sharma, chief executive of TeamLease Digital, said the five largest IT companies reduced their combined headcount by nearly 7,000 in FY26, compared with an addition of around 12,700 in FY25. At the same time, industry-wide employment grew by about 1.35 lakh in FY26, only marginally above the addition of 1.33 lakh in FY25.
Those numbers show why the current cycle cannot be described simply as either expansion or contraction. The wider industry is still adding workers, but the pace is restrained and the distribution of opportunities is changing. Revenue may be growing, yet headcount is growing more slowly. Attrition is increasingly being backfilled selectively rather than on a one-for-one basis, with companies redistributing work, relying on internal talent or recruiting for a different capability.
This is different from the large-scale workforce corrections associated with 2022 and 2023. Anil Ethanur, co-founder of Xpheno, said the high-volume layoffs during that period were used to reduce excess capacity created during post-pandemic hiring and to respond to margin pressure in a slower market. Large technology companies have since completed much of that downsizing and have kept headcounts under control through lower hiring over more than two years, he said.
The current phase is consequently less visible but potentially more structural. Fewer companies may be announcing blanket layoffs, while targeted exits, tighter performance management, thinner middle layers and slower replacement of departing employees continue. Workforce optimisation is increasingly framed around efficiency rather than only cost reduction. Performance Improvement Plans can provide a structured opportunity to improve performance, Ethanur said, although a failed plan can still result in an exit and should not be treated as a softer substitute for layoffs.
The pressure is concentrated in the routine parts of technology work. Milind Shah, managing director of Randstad Digital India, identified some entry-level coding, routine testing, basic application maintenance and repetitive data-processing activities as particularly exposed to artificial intelligence. The common feature is not that the work is labelled “technology”, but that it is repetitive, rules-based and highly standardised.
This changes the meaning of entry-level employment. Traditionally, low-complexity engineering output was often assigned to junior and mid-level workers, allowing them to build experience inside large organisations. If automation absorbs more of that work, the first rung of the career ladder can become narrower even when demand for advanced technical work rises. Ethanur said junior and mid-level engineers were among the most vulnerable, while low-complexity functions performed by more experienced workers could also face exposure.
The report also describes a challenge across the hierarchy. Ethanur said enterprise AI platforms have the potential to affect low, mid and high cognitive layers of the workforce, estimating that 40% to 60% of technology employees engaged in low- and mid-cognitive functions could be affected as enterprise AI matures. This is presented as an expert assessment, not as a measured industry-wide outcome. Its significance lies in the possibility that experience alone may no longer protect workers whose tasks remain standardised.
That makes the transition different from earlier technology disruptions. The issue is not only whether a role disappears. It is whether the content of an existing role changes faster than an employee can acquire the required skills. A software professional may remain employed but face a different expectation around automation, data, client interaction, domain knowledge or the use of AI tools. The risk is therefore also one of occupational downgrading or stalled progression, not only outright job loss.
Attrition patterns reinforce this interpretation. TeamLease Digital places current attrition at large IT services companies broadly in the 13% to 14% range, significantly below the post-pandemic peak. Yet the destination of workers is becoming more important than the aggregate rate. Employees with AI, cloud, cybersecurity and other specialised capabilities continue to have mobility, while routine roles and some middle-management layers face restructuring.
The reasons employees leave also point to a changing employment contract. Randstad research cited in the report says 40% of IT workers globally have left a job because of a lack of flexibility, while 43% have left because of limited career advancement. Nearly one in four technology professionals globally have left because their employer did not provide adequate upskilling opportunities, and 52% are pursuing training independently. The report says a similar emphasis on advancement is visible in India, though it does not provide a separate Indian percentage for each measure.
This has an urban dimension because access to future-relevant work is uneven. Workers in large technology centres may have more employers, training providers and specialised networks around them, while those outside established clusters may have fewer ways to move into AI, cloud or cybersecurity roles. However, the supplied evidence does not establish how these effects vary by city, income group or gender. What it does show is that the value of location-based access to employers is being joined by the value of capability-based access to work.
The demand side is becoming clearer. Hiring is moving from volume-led recruitment towards capability-led recruitment, with demand concentrated around AI and machine learning, cloud, data science and analytics, cybersecurity and automation. Professionals who combine AI fluency with technical expertise are expected to be more valuable, especially when they also bring problem-solving ability, domain knowledge and client-facing skills.
This combination is important because the industry’s transformation is not being described as a replacement of all human work by machines. Companies still need workers who can understand clients, translate business requirements, manage risk and apply technology within specific domains. What is changing is the balance between execution and judgement. Routine execution is more exposed when it can be standardised, while work requiring context and cross-functional understanding becomes more defensible.
The institutional response is currently centred on selective hiring and reskilling rather than an announced broad employment programme. Randstad’s Workmonitor study, as cited in the report, says 95% of employers globally expect their organisations to grow in 2026, even as hiring becomes more focused on specific capabilities. Nearly nine in ten Indian employees consider reskilling important. These figures suggest that growth expectations and workforce anxiety are coexisting rather than contradicting each other.
For companies, this creates a planning problem. Leaving vacancies open can improve productivity if work is redesigned effectively, but it also places greater pressure on existing teams and can reduce opportunities for junior employees to learn through routine assignments. Selective hiring can produce a more specialised workforce, while internal reskilling can preserve institutional knowledge. The report does not establish which approach produces better employment outcomes, but it makes clear that replacement is no longer automatic.
For workers, the central question has shifted from whether the technology sector is hiring in aggregate to which capabilities employers are willing to fund. Career mobility is increasingly tied to access to training, flexibility and opportunities to work on higher-value assignments. This does not mean every employee must become an AI specialist. It does mean that roles built primarily around repeatable tasks face greater pressure as employers seek productivity gains.
The available evidence confirms a selective restructuring of Indian technology employment rather than a broad-based collapse. Large layoffs have eased, industry employment has continued to grow and specialised skills remain in demand. At the same time, mid-career layers, routine functions and some entry-level work face greater uncertainty, while backfilling is becoming more cautious.
The next phase will be defined by how quickly organisations convert training into actual internal mobility, whether entry-level workers can still gain experience, and whether productivity-led growth creates enough new roles to offset the work being redesigned. The report establishes the direction of change, but not its eventual scale across companies or cities. That remains the key development to monitor as Indian IT moves from adding more people to extracting more value from different skills.