India’s GCC Hiring Hit a Record 510,000 Jobs in 2026, and the Work Looks Nothing Like Before

The GCC hiring number that is easy to misread

GCC hiring in India crossed a genuine milestone in 2026: foundit’s Insights Tracker, cited by Business Today in early July, projects 510,452 jobs across Global Capability Centres this year, a 3.4x increase since 2021. The country now hosts close to 2,120 such centres, a count that lines up closely with Zinnov and Nasscom’s own FY26 tally of 2,117. First-half hiring alone came to 227,991 additions, up 11% year-on-year, so the pace is not levelling off.

Most coverage of that milestone stops at scale: more centres, more jobs, another record year for India’s back-office economy. That framing was accurate for the GCC wave of the 2010s, when a Bengaluru or Hyderabad centre existed mainly to do the same finance, HR or IT-support work the parent company did in Ohio or Frankfurt, at a fraction of the cost. It is a poor description of what is being hired for now.

From cost arbitrage to the AI control layer

The tell is in the skill mix. Nearly two in three new GCC roles created in 2026, 64% by foundit’s count, now call for AI, data science or intelligent automation expertise. AI and data-science hiring is expanding at 38% year-on-year, faster than any other function GCCs track, and it is pulling ahead of the traditional IT-support and finance-operations roles that used to define the sector.

Broken down by sector, technology and software work still takes the largest single slice of GCC hiring at 31%, but AI, data science and analytics now sit at 18% and engineering or product R&D at 16%. Add those three together and AI-adjacent work already outweighs every legacy back-office function combined, including BFSI operations (9%), finance and accounting (6%) and HR or administration (4%). The centres being built now are not cost sinks; they are where a growing share of the parent company’s actual AI product work happens.

The job titles reflect it too. MLOps specialists, applied AI engineers, model evaluation leads and prompt or context engineers now show up on GCC job boards in categories that barely existed five years ago, sitting alongside the finance and HR roles a GCC was originally built around rather than replacing them outright.

Bengaluru still wins the count, not the growth rate

Bengaluru remains the anchor of India’s GCC map, with 30% of all hiring in 2026 and a 10% year-on-year rise of its own. But that is no longer the fastest-growing number in the dataset. Hyderabad, at 15% share, is growing 15% year-on-year; Pune (12% share) is up 11%; Mumbai (11%) is up 8%; Chennai holds 9% and Delhi NCR 8%. Every established metro is still adding headcount. None of them is where the acceleration is happening.

Tier-2 cities, taken together, hold a comparatively modest 15% share of GCC hiring, but that share is expanding at 23% year-on-year, nearly double the pace of Bengaluru and close to double most of the other metros individually. A city that barely registered on a GCC site-selection shortlist five years ago is now the fastest-moving line on the chart.

What Coimbatore and Indore actually look like up close

Coimbatore is the clearest individual case. Zinnov’s analysis of tier-2 GCC hubs puts the city’s five-year CAGR in new centre setups at 21%, well above the national base rate. It now hosts more than 50 GCCs employing upwards of 11,000 professionals, and roughly 60% of that workforce sits in engineering R&D roles rather than the transactional support functions that defined earlier tier-2 experiments. That is not a call-centre expansion story; it is an engineering one.

Indore tells a cost story more than a talent-depth one. Recruiter benchmarking from Plugscale’s 2026 GCC compensation analysis puts Indore’s total cost index 30-40% below Bengaluru’s, and Madhya Pradesh’s state IT and ITeS incentive schemes layer capital subsidies, stamp-duty waivers and power-tariff rebates on top of the wage gap itself. Kochi, Jaipur, Bhubaneswar and Visakhapatnam show up in the same tracker as smaller but consistently growing hubs, each pitching some combination of the two arguments: either the engineering bench is deep enough to matter, or the total cost of running a centre there is meaningfully lower.

The number that should worry Bengaluru incumbents more than either of those is attrition. Tier-2 GCC hubs report 8-12% turnover against 18-22% in tier-1 metros, a gap wide enough to change the economics of any team running production AI systems, where the person who built the evaluation harness is worth more the longer they stay on it.

The talent pipeline underneath this is shifting as well. Zinnov has previously pointed out that roughly 60% of India’s graduate population sits outside the metro belt altogether, which is as much a reason tier-2 GCCs are opening as any tax incentive: the engineering colleges feeding Coimbatore, Indore and Kochi are producing graduates who once had to relocate to Bengaluru or Hyderabad for a GCC-grade job and increasingly do not have to.

The pay math is compressing towards the tier-2 case

Compensation is where the arbitrage story gets complicated. Plugscale’s 2026 GCC benchmark puts Bengaluru AI-engineer pay at ₹28-45 lakh a year for mid-level roles, ₹48-72 lakh at senior level and ₹78-120 lakh at lead or principal grade, a 40-60% premium over backend engineers at equivalent seniority. Those are not back-office wages by any definition; they sit within striking distance of what a mid-career engineer earns at a well-funded product company in the same city.

Remote AI hires working out of tier-2 cities like Coimbatore or Indore for US-headquartered firms are reported to command the rupee equivalent of ₹60-80 lakh a year, while still carrying tier-2 living costs. Put the two figures side by side and the incentive for a global company becomes obvious: hire the specific engineer wherever they sit, rather than building headcount in the city with the highest average cost and hoping seniority follows.

There is a structural tell buried in the age mix too. Engineers with 4-10 years of experience make up 56% of 2026 GCC hiring, the largest single band, but entry-level hiring (0-3 years) is growing fastest of any experience group at 18% year-on-year even though it is a smaller 30% share overall. Companies are hiring junior AI talent to grow into the senior roles in-house, not assuming they can buy seasoned applied-AI engineers off the shelf at scale.

What this changes for whoever picks the next site

For a VP of engineering or a chief people officer sitting down to plan a 2027 GCC footprint, the practical read is narrower than the trend pieces suggest. Skill depth, not headcount, is now the site-selection metric that matters for any AI-linked role: a city needs a genuine bench of applied-AI and data engineers, not just a lower blended salary line, or the centre will spend its first eighteen months hiring rather than shipping.

Attrition math favours tier-2 hubs for exactly the kind of work that compounds, production AI systems, evaluation pipelines, MLOps, where institutional memory is worth real money. Bengaluru still wins for architect-level and staff-plus hiring, where the bench of people who have done the job three times before is deeper than anywhere else in the country. Increasingly, the answer is not one city; it is a split between a senior anchor centre in a metro and a growing junior-to-mid engineering base in a tier-2 hub feeding into it.

In practice that means treating a GCC location decision less like a real-estate exercise and more like a build-versus-buy call on talent. A pilot cohort of ten to twenty engineers run for two full quarters before signing a lease tells a company more about a city’s actual bench than any consultancy deck. Attrition should be benchmarked after the first year of a cohort, not the first quarter, since tier-2 retention gaps often widen only once the initial group has been through a full appraisal cycle. And site selection should be priced against the specific AI-linked role the parent company needs shipped, not the median GCC role, because that is the one it actually cares about.

Where the growth story could still break

The forecast attached to all of this, a workforce of 3.4-3.5 million GCC professionals by the end of the decade if current trends hold, carries a genuine conditional. Hiring intensity by sector is uneven: BFSI and fintech GCCs are hiring at 1.17 times the base rate, technology and software at 1.13 times, while energy, chemicals and other legacy sectors sit at 0.50 times. This is a boom concentrated in AI-linked and finance-linked demand, not a rising tide across every industry that runs a GCC.

That concentration is also the risk. Older GCC hiring cycles tracked global labour-cost arbitrage, a slow-moving variable that rarely reversed inside a single budget year. A hiring cycle now tied to AI product roadmaps and parent-company AI capital spending is tied to a far more volatile input. If global HQ budgets for AI headcount tighten in 2027, the GCCs most exposed will not be the legacy back-office ones; they will be the AI-first centres that grew fastest precisely because they were AI-first.

There is a second, quieter risk in the entry-level surge itself. Hiring graduates faster than a GCC can season them into senior applied-AI roles only pays off if the intervening years of mentorship and project depth actually happen, and a centre under pressure to show quick AI headcount can be tempted to skip that step. The gap between hiring for AI skills and actually operating at senior AI depth is where several of the coming years’ disappointing GCC case studies will likely come from.

The next wave will not look like the last one

The GCCs opening in 2027 are being built on a different premise than the ones that opened in 2015: not lower cost for the same work, but a real claim on the work itself. The cities picking up that next wave, Coimbatore’s engineering bench, Indore’s incentive stack, Kochi and Bhubaneswar’s early bets, are competing on talent depth and retention as much as on the rupee-dollar spread that built the industry in the first place. Whether India’s GCC workforce actually reaches 3.5 million by 2030 depends less on how many centres open next and more on whether the AI-linked demand driving 2026’s hiring holds its pace once the first wave of parent-company budgets gets tested.