Verdict: AI is not yet destroying India's IT jobs in mass layoffs — it is freezing new hiring by diverting enterprise budgets and flattening the software demand curve. Zoho founder Sridhar Vembu's August 2026 warning that "we have not laid off people, but we are not creating new jobs" is the more uncomfortable signal: the contraction is structural, not cyclical, and it hits the companies least likely to use headcount as a lever. The data backs him.
Last verified: 2026-08-03
- The IT sector, including Zoho, has stopped net new job creation; layoffs and hiring are separate phenomena.
- TCS cut ~12,200 jobs in FY26 while its chairman called AI "the biggest opportunity" — the two stories conflict.
- Entry-level IT roles fell 20–25% in 2025 (EY); India's tech workforce growth slowed to 2.3% (NASSCOM).
- Enterprise IT budgets are reallocating from people to AI infrastructure and data centers.
- India's demographic dividend peaks by 2030 — the window to absorb youth into IT is closing.
- Pricing/figures change often — last checked 2026-08-03.
What did Sridhar Vembu actually say about AI and IT jobs?
On August 2, 2026, Zoho Corporation founder Sridhar Vembu posted on X that India's IT industry — including his own company — has "not created many jobs in recent years," even though it has avoided mass layoffs. The cause, he wrote, is that "the money that would have gone to new employees is now going to AI and data centre costs, the latter due to the steep rise in server and memory prices" (Rediff/PTI, Aug 2 2026; India Today, Aug 3 2026).
The distinction he drew matters more than the headline suggests: Zoho has not cut a single job, but it is also not creating new ones. A hiring freeze at a company deliberately engineered to avoid layoffs is a different signal than a layoff at a firm under quarterly pressure. Vembu framed the challenge as national: "The real issue facing our nation is how to create jobs for our massive cohort of youth in this very uncertain global landscape" (Indian Express, Aug 2026).
Why is a hiring freeze at Zoho more alarming than layoffs elsewhere?
Zoho is the company least likely to treat headcount as a financial lever. It is privately held, profitable, bootstrapped (no external investors since founding in 1996), and headquartered in rural Tenkasi, Tamil Nadu — a deliberate choice to create jobs outside Bengaluru's metro corridor (Wikipedia: Sridhar Vembu; Global Indian profile). Vembu moved from Silicon Valley to Mathalamparai near Tenkasi in 2019 specifically to prove world-class products could be built from villages, and Zoho's Tenkasi office has grown from 6 employees in 2011 to over 500 today.
When the company built to avoid the headcount-as-lever trade-off publicly states it cannot create new roles, the signal is about the economics of software work itself — not Zoho's choices. Contrast this with Tata Consultancy Services (TCS), India's largest IT employer, which cut approximately 12,200 positions in FY26 (roughly 2% of its 613,000-strong workforce) — its largest-ever reduction (Indian Express, Jul 27 2025; Final Round AI summary). TCS is publicly traded and answers to shareholders; Zoho answers to no one. Yet both are shrinking net hiring.
Does the world actually need more software?
Vembu posed a question few in Indian IT ask publicly: "While AI allows us to produce more software quicker, does the over-saturated global software market need a lot more software?" (Rediff/PTI). His analogy is the mature commodity market: when an industry saturates, growth slows and competition shifts from volume to quality, reliability, and brand. Faster code production under that logic does not create more jobs — it means fewer people produce the same output.
This is the structural argument, and it is harder to rebut than the simpler "AI replaces workers" narrative. If enterprise customers already have enough software, AI-driven productivity gains compress the labor needed to maintain and extend it, even without a single layoff. The hiring freeze is the displacement, just delivered quietly and deferred to the next cohort of graduates who never get hired.
How are enterprise IT budgets shifting toward AI?
Two budget movements are squeezing traditional IT hiring simultaneously:
Server and memory cost inflation. The AI buildout has driven up data center infrastructure costs — GPUs, HBM memory, cooling, and power. Vembu explicitly cited "the steep rise in server and memory prices" as diverting funds that would have gone to new employees (Rediff). This is the same cost pressure we analyzed in the AI spending paradox driving down tech stocks and the rollback of data center tax breaks.
Customer budget reallocation. Vembu noted that enterprise customers are "redirecting their own IT spending towards AI" (India Today). This means the pie isn't growing for traditional IT services — it is being repartitioned. He also expressed doubt that the AI companies spending heavily on capex "will achieve the massive profits they need to justify all the capex," suggesting the reallocation may be speculative.
For an IT services industry that sells person-hours to enterprises, a customer who reallocates from "more software" to "AI infrastructure" is a customer buying fewer billable hours. That translates directly into fewer hires.
What does the hiring data actually show?
The Vembu warning is not isolated — the numbers have been deteriorating across the sector:
| Metric | Figure | Source |
|---|---|---|
| Entry-level IT roles decline (2025) | 20–25% | EY analysis, via Mint |
| India tech workforce growth (FY26) | 2.3% (slowed) | NASSCOM, via Mint |
| Tech job openings, Jan 2026 vs Jan 2025 | −24% | Xpheno, via Economic Times |
| Entry-level share of tech hiring (2024→2025) | 28% → 15% | Metaintro |
| TCS FY26 workforce reduction | ~12,200 jobs (~2%) | Indian Express |
| TCS AI services revenue (annualized) | $2.3–2.5B | TCS Annual Report 2025-26, via Economic Times |
The pattern is consistent: mass fresher hiring (the traditional on-ramp that absorbed millions per year) has contracted sharply, while spending and hiring in AI-adjacent specializations continues. NASSCOM and Deloitte project India's AI talent pool will need to reach roughly 1.25 million people by 2027 — a large number, but a fraction of the 7–10 million young workers entering the labor force annually (Metaintro; pwOnlyIAS, May 2026).
How do Vembu and Chandrasekaran's positions conflict?
The contrast between the two most recognizable voices in Indian tech is the article's central tension:
N. Chandrasekaran, TCS chairman, told shareholders at the FY26 AGM (June 2025) that AI is "the biggest opportunity in the company's history" with "no downsizing plan," and that fears of software firms "being left behind" by AI have "not been our experience" (TCS Annual Report 2025-26, Chairman's letter; News18). TCS reported annualized AI services revenue of $2.3 billion and new-age services (cloud, data, cybersecurity) revenue of $11.5 billion (Economic Times).
Yet TCS cut ~12,200 jobs in the same fiscal year — its largest-ever workforce reduction — citing skill mismatches, bench policy changes, and macro uncertainty (Indian Express).
Two leaders, the same slowdown, opposite public narratives. One describes what customers are doing with budgets; the other describes what investors need to hear. Vembu — accountable to no one but himself and his employees — is free to say the quiet part out loud.
Can manufacturing absorb the workers IT no longer hires?
Vembu's answer was blunt: no. "Extensive automation means large-scale production now creates comparatively few jobs," he wrote (Rediff). Manufacturing automation makes goods cheaper but employs fewer people per unit of output. If both software and large-scale manufacturing shed labor while producing more for less, the harder question becomes: how do people earn enough to buy what the machines make?
This is the income-distribution problem Vembu flagged as structural rather than temporary. He suggested Universal Basic Income (UBI) — already visible in India in the form of welfare "freebies" — may see growing political pressure as the youth-employment challenge persists. The budget math is formidable: a minimal UBI at India's poverty line (₹7,620 per person annually) would cost roughly 5% of GDP (Drishti IAS analysis).
What is India's demographic timeline?
India's demographic dividend — the working-age population share — will begin declining after 2030, per the Azim Premji University State of Working India 2026 report (APU, 2026; Mint analysis). Graduate unemployment among 15-to-25-year-olds stands near 40%, and only a small share of graduates secure stable salaried jobs within a year of finishing their degrees. The traditional IT on-ramp — the thing Vembu says is now frozen — was one of the few reliable bridges from education to salaried employment that scaled to millions.
The window is roughly four years. If AI freezes IT hiring through that window, the demographic dividend risks becoming a demographic liability — a large, educated, underemployed young population with rising expectations and shrinking on-ramps. India's broader tech hub strategy, including the Global Capability Centre boom, is one partial substitute; GCCs grew demand 13% month-on-month in early 2026 (Economic Times/Xpheno), but at ~17,000 openings it is an order of magnitude smaller than the IT services intake.
What should IT workers and graduates actually do?
The practical picture from the data, not the rhetoric:
- Move up the value stack. The roles contracting fastest are basic software testing, data entry, and routine maintenance — work that is repetitive and rules-based. The roles holding steady are senior architecture, client-facing relationship management, and specialized skills in cloud, data engineering, and AI tooling (Final Round AI; Mint).
- Build AI fluency, not just AI awareness. Employers want candidates who can demonstrate judgment about what AI should vs. shouldn't do — reviewing and refining AI-generated output, not just producing code. A portfolio with a real shipped project beats a degree alone (Metaintro).
- Look beyond IT services. Global Capability Centres, AI engineering, cybersecurity, and data science are the clusters still growing. NASSCOM/Deloitte project a 1.25M-person AI talent gap by 2027 — but these are rarely framed as entry-level, so upskilling before applying matters.
- Don't wait for the cycle to turn. Vembu's point is that this is not a downturn that recovers; the structure has changed. Treating it as a temporary freeze (the way prior slowdowns were temporary) is the risk. If you want to go independent instead, see how to start a one-person AI consulting business or build a no-code AI agent team as a solopreneur.
What this means for you
If you are an IT worker in India (or anywhere the same pattern is emerging — the US saw programmer employment drop 27.5% between 2023 and 2025 per BLS data cited by IEEE Spectrum), the actionable read is: the commodity layer of software work is contracting and will not come back. The defensible positions are judgment-heavy, system-level, and AI-augmented. The companies that still describe AI as "the biggest opportunity" are describing their opportunity in selling AI services, not your opportunity in keeping the job you had. Plan accordingly — the hiring freeze is the leading indicator, and it has already arrived.
FAQ
Q: Is AI causing layoffs in India's IT sector? A: Not yet at scale — the major Indian IT layoffs (TCS's ~12,200 cuts in FY26) were attributed to skill mismatches, bench policy changes, and macro uncertainty, not direct AI replacement. The bigger signal is the hiring freeze: AI is freezing net new job creation by diverting budgets and compressing software demand, not destroying existing roles.
Q: What did Sridhar Vembu say about AI and jobs? A: On August 2, 2026, Vembu posted on X that the IT industry, including Zoho, had "not created many jobs in recent years" because money that would have gone to new employees is now going to AI and data center costs (driven by server and memory price inflation). He has not laid off staff but is not creating new jobs — a structural hiring freeze.
Q: How many jobs did TCS cut in FY26? A: Tata Consultancy Services announced approximately 12,200 position cuts (about 2% of its 613,000-strong workforce) for fiscal year 2025-26 — its largest-ever workforce reduction. CEO K. Krithivasan and chairman N. Chandrasekaran attributed it to skill mismatches and restructuring, not AI directly, even as TCS reported $2.3B in annualized AI services revenue.
Q: Is India's demographic dividend at risk from AI? A: The Azim Premji University State of Working India 2026 report says India's working-age population share will begin declining after 2030 — a roughly four-year window. With ~7–10 million young workers entering the labor force annually and entry-level IT roles down 20–25%, the traditional IT on-ramp that absorbed millions is narrowing at exactly the wrong time.
Q: Can manufacturing absorb workers displaced from IT? A: Vembu argues no — extensive automation means large-scale manufacturing creates comparatively few jobs per unit of output. Both software and manufacturing are shedding labor while producing more, which raises the income-distribution question (how people earn enough to buy what machines produce) more than the job-creation question.
Q: What skills are still in demand in India's IT sector? A: AI engineering, data science, MLOps, cloud architecture, and cybersecurity. NASSCOM and Deloitte project India will need an AI talent pool of ~1.25 million people by 2027. But these are rarely framed as true entry-level roles — employers want demonstrated portfolio work and judgment, not just a degree.
Discussion
0 comments