AI and India's IT Freshers: What the Data Shows vs the "6 Million" Claim

A headline circulating today says AI is slowing entry-level hiring for India's 6 million IT workers. The direction of the story is supported by data. The 6 million figure, and the idea that AI is the single cause, are not. This post sorts the claim from the evidence using company statements and staffing data, in the same claimed-versus-verified format we use for other contested numbers.

This is a follow-up to earlier worker-sentiment coverage like our look at India's AI workers and layoff fears in a Blind survey, and a companion to the UK evidence in UK employers cutting entry-level jobs.

Claimed vs verified

What do the company numbers say?

Outlook Business, drawing on Mint's analysis of annual reports and Xpheno staffing data, reports that some of India Inc's largest employers are recording their lowest fresher-hiring numbers in years. Key figures from that piece:

• Xpheno: demand for entry-level tech professionals fell to 10,000 openings from 13,000 in May 2026, a 44% year-on-year decline.
• Infosys: employees aged 30 and below were about 50.7% of its 3.28 lakh workforce at the end of FY26, down from 53% in FY25 and the lowest in 15 years. Until FY18 that group was at least two-thirds. Management still indicated at least 20,000 freshers in FY26, a target it reportedly met.
• TCS: TCS ended FY26 with 5.84 lakh employees after a July layoff of around 12,000.
• Reliance: its FY26 annual report says "1 lakh+" new hires against over 1.9 lakh in FY25, suggesting a cut of close to 90,000, even as overall headcount grew with retail.

Wipro's numbers come from its own earnings call. CHRO Saurabh Govil said campus recruitment was muted in the October to December quarter with "only about 400-odd freshers" joining, and that against a stated 10,000 the company would end between 7,500 and 8,000. Wipro's year-to-date fresher intake was still above 5,000, and the company added 6,529 employees in the quarter mostly through the Harman DTS acquisition and contract redeployment.

Is AI the reason?

Here the sources part ways. Anand Rathi's Sushovon Nayak told Mint that automation tools are reducing the need for young graduates unless they have specialised skills, and that firms increasingly prefer employees aged 30 to 40 who can be upskilled. Everest Group's Peter Bendor-Samuel pointed to a different pressure: with new AI talent in demand, services firms must hire higher-calibre people who do not see them as a first choice.

Recruiters quoted by Mint gave two broader reasons for fewer new hires: redeployment driven by AI adoption, and a sluggish job market reducing voluntary exits. At a Reuters Summit, executives from Novo Nordisk, Daimler Truck and Epsilon said AI is raising productivity in India while headcount stays flat or "conservative."

Our reading: AI is one factor among several, and the data cannot isolate its share. A hiring freeze during a weak demand cycle looks identical in the numbers to AI-driven substitution. The honest way to test the AI hypothesis is to look at which roles shrink fastest (routine testing, support, junior coding) and whether output per head rises, which these sources do not measure. Similar caution applied when we checked the World Bank's jobs tsunami claim and Zuckerberg's claim about an abundance of jobs.

Why do the fresher targets conflict?

Fresher-hiring figures in circulation often conflict, and we have left out totals we could not trace to a company statement or a primary release. A company earnings call or annual report is the check.

The practical lesson is that "fresher hiring" can mean campus offers made, offers accepted, people who actually joined, or open job postings. Xpheno's openings count measures demand posted by employers; Wipro's 400 counts joiners in a single quarter; Infosys's 20,000 is a full-year target. Mixing those in one sentence is how a story ends up with a confident number that no single source supports.

How to read the next headline

1. Check the unit. Is it openings, offers, joiners, or headcount? Are the figures for one company or the whole sector?
2. Check the period. The Outlook Business piece combines FY25, FY26 and mid-2026 monthly data; a year-on-year change in one month is noisy.
3. Separate correlation from cause. If hiring falls while profits are flat and AI tools are being rolled out, AI is a suspect, not a verdict. Look for statements from the firms about what changed in project staffing.
4. Look for the contrary evidence. Infosys reportedly still met a 20,000 fresher target, and Wipro's year-to-date fresher intake stayed above 5,000, so the picture is a squeeze, not a collapse, at least for the largest firms.
5. Ask what is being replaced. The strongest AI argument is about task mix: if routine work done by juniors is automated, firms need fewer juniors per project, which shows up first as lower fresher targets rather than layoffs.

Where does the 6 million number come from?

We searched for it and could not trace it to a primary source. Different definitions give different totals, and the verified company figures we found are far smaller: TCS at 5.84 lakh and Infosys at 3.28 lakh. We are not repeating 6 million as fact. If you have the original report, check whether it counts the whole IT-BPM sector, includes contractors, or refers only to a subset. Industry body Nasscom and company filings are the places to verify aggregate numbers.

What are firms doing instead?

Wipro says it created 50 Centres of Excellence with universities, building curricula in AI, cybersecurity and data and hiring from them, according to the earnings-call coverage above.

For a student, the pattern is consistent: general entry-level seats are shrinking while seats attached to AI, data and security skills are being created, often through direct university pipelines.

What can freshers do?

1. Show work, not only degrees. Projects built with AI coding tools, with tests and writeups, are evidence of the skill employers say they want. Our loop engineering career guide for students is a starting point.
2. Target firms with campus pipelines in your area, since that is where Wipro-style hiring is happening.
3. Learn to verify AI output. Reviewing and testing is the part of the job that stays human; see how the verification tax shapes productivity claims.
4. Watch company disclosures (quarterly earnings calls, annual reports) instead of aggregate headline numbers.

What would change our assessment?

Three pieces of evidence would sharpen the picture. First, company disclosures that tie specific project types to reduced junior staffing, which would move AI from suspect to documented cause. Second, a longer Xpheno series showing whether the 44% drop persists once seasonal hiring cycles are accounted for. Third, a clear definition behind the 6 million figure, so the sector total can be compared with firm-level headcounts. Until then, the defensible statement is narrower than the headline: entry-level hiring at major Indian employers is down, AI is plausibly part of why, and no source we read proves its share.

Why does this matter beyond the headline?

Entry-level hiring is how large services firms have historically built their talent pipeline. Companies like TCS, Infosys and Wipro recruit thousands of graduates each year, train them in-house, and promote the strongest into delivery and management roles. If the intake narrows for several years, the effect shows up later as a thinner middle layer, not only as fewer first jobs today. That is a reason to track fresher numbers closely even when the cause is unsettled.

It also matters for graduates and colleges. Engineering campuses in India have long treated mass recruiters as the default outcome for a large share of students. A shift toward fewer, more specialised hires changes what a curriculum should teach, and it raises the value of internships, open-source contributions and portfolio projects that signal skill beyond marks. The Wipro Centres of Excellence model, where a firm shapes the course and then hires from it, is one version of how that adjustment could work.

Finally, it matters for how we read AI-and-jobs claims generally. Several outlets have published confident totals on AI job losses, and few of them state the unit, period or definition. Treating each number as a claim to be checked, as the table above does, is more useful than choosing a side in the optimism-versus-alarm debate. A narrower statement that survives scrutiny beats a dramatic one that cannot be traced.

What is still unknown

We did not find a fresh October 2026 report behind today's headline, nor a source for the 6 million workers figure. Company numbers cited are from FY26 filings and calls, and the Outlook Business piece mixes data from several months. The next data points to watch are second-quarter FY27 results from the large IT firms and Xpheno's next monthly update.

Related reading

• India's AI workers and layoff fears: Blind survey
• UK employers cut entry-level jobs
• World Bank East Asia update: the jobs tsunami claim
• Zuckerberg's abundance of jobs claim, checked
• Loop engineering career guide for students
• The verification tax on AI output

Figures reflect the cited reports and company statements as of October 13, 2026 and may change.