Vantage
At the Zinnov Confluence that I attended in Bengaluru last month, one small change created more impact than most of the keynotes. Zinnov introduced a new AI Excellence award, and the criterion was impact, not adoption.
“Impact, not adoption.”
I think that’s the right bar, because India's GCCs don't lack adoption. According to the Nasscom-Zinnov GCC Landscape Report 2026, more than 1,200 India GCCs have AI/ML capabilities, with over 250,000 AI/ML professionals. Nearly half of the GCCs set up since FY2021 were AI-first from day one.
Yet, a Dell/Zinnov report found that nearly 70% of India's GCCs remain stuck at the AI pilot stage.
In my experience, talent that drives AI adoption is not the talent that delivers AI impact, and most GCCs are still hiring for adoption.
Adoption talent can use and demonstrate AI tools: prompt skills, model familiarity, quick prototypes. Impact talent can make AI work inside a real business deliverable. These are people who understand the enterprise's data and systems, can build AI into everyday workflows, and can set realistic expectations with leadership.
Adoption talent gets you pilots. Impact talent gets you results.
1. Pilots Stall Because Organizations Aren't Ready
S&P Global Market Intelligence surveyed more than 1,000 enterprises in N America and Europe. It found that the share of companies abandoning most of their AI initiatives rose to 42% in 2025, from 17% a year earlier, and that the average organization scrapped 46% of its AI proofs of concept before they reached production. Gartner's latest survey of 782 infrastructure and operations leaders found that only 28% of AI use cases fully met ROI expectations, and 20% failed outright.
In the same Gartner survey, the most common cause of failure besides over-ambition (organizations expected too much, too fast), was the skills gaps and poor data quality. The report concluded that ROI depends less on how sophisticated the model is and more on how well AI is integrated into real operations.
In my view, each of these causes comes back to talent. Unrealistic scoping, poor data and weak integration aren't fixed by hiring more generalist AI talent. They're fixed by people who understand how the business actually works.
ROI depends less on how sophisticated the model is and more on how well AI is integrated into real operations.
2. Adoption Skills Are Becoming a Commodity
When AI itself becomes easy to access, so do the skills needed to adopt it. That is exactly what is happening now. OpenAI began rolling out GPT-6 Astra through its enterprise plans. Nvidia agreed to acquire Hugging Face, the leading platform for open-source AI models. The most powerful AI is now available to every enterprise, including your competitors.
When something becomes a utility, value moves to what you do with it. When factories first got electricity, most simply replaced their steam engine with a large electric motor and changed nothing else. The real productivity gains came years later, when manufacturers redesigned the factory floor around the new power source.
Many GCCs today are swapping the steam engine for a motor: bolting AI on to existing processes and hoping for a productivity jump. And from what I have seen within the GCC ecosystem, many are paying premiums for skills that mainly connect off-the-shelf models to business processes, just as those skills are becoming commodities. Another Zinnov/Indiaspora study found that more than half of GCC work in India faces displacement pressure or commoditization because of AI.
So the scarce skills are with people who understand your enterprise's data, systems and processes well enough to redesign work around AI.
When something becomes a utility, value moves to what you do with it.
3. The GCCs Delivering Impact Own the Foundation
JPMorgan Chase, as reported by NYPost last year, had more than 450 AI use cases in production and estimated up to $1.5 billion in annual value from AI, built on years of investment in data and platforms before Gen AI arrived. Walmart Global Tech's India teams built recognized as a Zinnov GCC Icon this year AI systems for supply chain, pricing and omnichannel retail that run across Walmart's global store network, with India-based leaders owning global product lines. TIAA Global Capabilities won this year's AI Excellence award and has now leased more than 250,000 sq ft in Pune to scale its AI, data and analytics teams.
Even Nvidia is pointing AI at its own operational data first, through a new partnership with Palantir. When everyone can rent the same models, advantage belongs to whoever owns the data and the integration. Pilots and vendors can come and go, but the GCC that holds the foundation becomes indispensable.
When everyone can rent the same models, advantage belongs to whoever owns the data and the integration.
What GCC Leaders and Boards Should Do
- Rewrite the hiring job description. For every AI specialist you hire, ask who will make their work succeed in production. Rebalance towards people with deep process knowledge.
- Fund foundations before pilots, and set realistic expectations up front.
- Measure AI in live use, not pilots launched. Pilot counts measure activity. Value in production measures capability.
- Rebuild the talent pipeline and design new career paths as AI slowly removes the junior work that used to develop senior engineers.
Three Questions Every Board Should Ask Its GCC
- What share of our AI investment goes into foundations: data, integration and skills versus pilots?
- How many AI initiatives are in live use, and what value have they delivered against realistic targets?
- Who understands our core data and processes well enough to make AI work at scale, and how are we developing the next generation of them?
The Pressure Is Building
Bloomberg reported recently that Goldman Sachs projects that AI infrastructure spending by the five largest US hyperscalers will rise to $1.2 trillion next year and that they would need about $300 billion a year in AI revenue to break even. That pressure for returns is likely to flow down to many enterprises and, in turn, their GCCs.
At home, the government is weighing a National Frontier AI & Compute Fund of up to INR 20,000 crore for GPUs, data centers and indigenous models. That builds the power station. It will be up to the GCCs to redesign the factory.
Adoption gets you nominated, but it's the impact that takes the award. And impact comes down to what skills you hire.
