In partnership with
Hello! 👋
It's Thursday, 7th May 2026. Welcome back to Bold Efforts!
So, a Global Capability Centre is a company's own office abroad, staffed by its own people rather than outsourced to a vendor. And this GCC success often gets written as an India success story. A country once described as a cheap labor market now runs AI infrastructure for JPMorgan, builds pricing algorithms for Walmart, and houses Goldman Sachs's second-largest global office. That is all true. What I keep thinking about is the question it leaves unanswered.
In 1985, Texas Instruments moved satellite communication equipment through Bangalore on bullock carts to connect its new R&D center with its headquarters in Dallas. TI was building India's first wholly owned multinational R&D center, hiring engineers to work on digital signal processors that shipped in products worldwide. Real engineers, doing real R&D.
But TI was the exception. The mass-market version of the India model had a different origin. Companies like TCS, Wipro, and HCL built something almost industrial in scale: fresh graduates hired by the thousands from campus recruitment drives, trained for a few months on sprawling campuses in Chennai and Mysore, then deployed to handle the IT needs of Western companies at a fraction of what local staff would cost. Y2K patches, helpdesk queues, basic software maintenance. The model was cheap capacity, deployed at scale.
Eventually Western companies asked a reasonable question: why pay the Indian IT company's margin? That is when the captive model took off. Goldman Sachs arrived in Bangalore in 2004 with roughly 300 people doing IT support. Today it has more than 9,000 in India, making it Goldman's second-largest global presence after New York. JPMorgan employs 55,000 people in India, one in five of its entire global workforce, and is building Asia's largest corporate campus in Mumbai for 30,000 more. Walmart's inventory forecasting models run out of India. McDonald's largest technology center outside the United States is in Hyderabad. There are now more than 1,700 of these centers across India. The industry calls them Global Capability Centers, or GCCs, unrelated to the Gulf Cooperation Council, and two to three new ones open every week.
This stopped being outsourcing a long time ago.
The work moved through four distinct phases over forty years: from Y2K coding to business processing, then to real engineering, and now to product development, AI infrastructure, and IP creation. The terminology shifted at each step, from captive center to Global In-house Center to Global Capability Center. Each renaming was an attempt to shed the old description. India was no longer just cheaper. India was capable.
Here is what that framing leaves out. GCCs are excellent for the shareholders of JPMorgan and Goldman: world-class work at a significant cost advantage. They are good for the engineers inside them: above-market salaries, serious work, a multinational brand on the resume. But the workers in New Jersey and Manchester whose jobs moved to Bangalore did not win. The Indian startup trying to hire a senior engineer against Goldman's salary offer does not win. The broader Indian tech ecosystem, which loses its most talented people to executing someone else's roadmap instead of building their own, does not win. Every model that creates winners creates losers. The GCC story usually counts only one side.
There is also a quieter cost. The most talented engineers in the country spend their careers executing someone else's vision. Take a senior developer at Goldman's Bangalore office. The work is excellent, the pay is real, and the systems they build move billions. The decision about what to build, the product roadmap, the company's direction over the next three years, was made in New York. India executes. The direction comes from somewhere else.
This sounds like a minor distinction until you spend a few years inside it. There is a ceiling in that arrangement, and it has nothing to do with ability.
The industry noticed this problem early. Its response, each time, was to rename the arrangement. Captive center became Global In-house Center. Global In-house Center became Global Capability Center. Each new name announced that the relationship had genuinely changed, that India was now a real partner rather than a delivery arm. The work did get more sophisticated. The question of who owns the vision stayed where it was.
You can rename the arrangement. You cannot rename the ceiling.
Two forces are now converging on this model at the same time. The first is AI: HSBC has announced plans to cut 20,000 operational roles globally, explicitly because of automation, and the work going first is exactly what the early GCC era was built on. The second is geopolitics: trade tensions, reshoring pressures, and security concerns are giving multinationals new reasons beyond cost to reconsider where they place sensitive work. Neither is fatal on its own. Together they put real pressure on a model that has never had to answer the question underneath it.
When execution gets automated and political winds shift, decisions are what remain. Who defines the product. Who owns the outcome. Who answers when the strategy turns out to be wrong.
This matters beyond the career stories of individual engineers. GCCs now account for roughly 25 percent of India's service exports and 40 percent of commercial real estate demand in its major cities. These are not vanity metrics. If the model contracts sharply, the ripple is large. India built 1.9 million careers on a foundation that assumed the world would keep wanting to locate execution work here. That assumption is now less certain than it has ever been.
The GCCs that last are the ones whose parent organizations are finally willing to move past execution. That means giving the people in Hyderabad the authority to originate the direction, own the outcome, and be wrong in front of headquarters. That is a different kind of trust than this industry has extended so far.
Whether the architecture changes this time, or gets a new name instead, is what the next decade will show.
Best,
Kartik
I write Bold Efforts every week to think clearly about where work and life are actually headed, not where headlines say they are. If you want these essays in your inbox, you can subscribe here.