The AI-First GCC: Why Offshore Banking Hubs Must Reinvent Work Now

Global Capability Centers (GCCs) in India have powered banking for the last 2 decades by delivering cost efficiency at scale through its deep, high-quality talent pools that offered abundant, technically strong and English-proficient talent across IT and operations.

In India alone, there are currently c1,800 GCCs supported by about 1.9 million professionals with a cost base of $64.6b according to Nasscom, with that number expected to touch 2,000 GCCs with 2.8m professionals costing $100b by 2030.

GCCs matured from the days of call centers to then IT services, and back and middle office operations with the level and capability of technology that existed then, but as technology evolved, GCCs started pushing into strategic hubs leading digital transformation, product engineering and analytics, and not just locally, but supporting speed to market for global programs as well.

But AI is rewriting the rules, various research indicates 2 in 3 jobs will be impacted, but equally new jobs requiring new skills will emerge as history has time and again shown us through every major disruption cycle.

The GCC Golden Era Is Ending

The current state of GCCs has its share of problems, which AI can easily fix including: -

• Fragmented, manual workflows - Many GCCs still run legacy, email- and spreadsheet-heavy workflows that create handoff delays, rework, and opaque ownership, especially across functions and locations.

  
  

•  Inconsistent quality and knowledge application - Quality varies by team, shift, and site because knowledge is embedded in people and local practices rather than codified in systems.

  
  

• Limited productivity visibility and reactive management - Many GCCs manage productivity via lagging, aggregate metrics, making it hard to detect bottlenecks, coaching needs, or value leaks at a granular level.

  
  

• Suboptimal customer experience in servicing - Traditional offshore models rely on script-based interactions, limited context, and linear IVR flows, often leading to long handle times and channel switching, and

  
  

• Attrition and wage inflation compounded by the “talent wall” – attrition rates average 12% - 15% with wage inflation for junior and mid-level roles standing at 10% - 25%, which are disproportionately higher for the incremental skillsets that employers pay for mainly due to wage wars especially in tier 1 cities. All of these reasons are eroding the cost arbitrage that made GCCs viable in the first place.

The Swivel Chair Trap

McKinsey’s ‘swivel chair phenomenon’ in banking operations captures this inefficiency vividly: long customer journeys are fragmented across grids of employees in different departments, each literally swiveling between systems and handoffs along a broken value stream before a single product or service is finally delivered to the customer. Analysts endlessly rekeyed data between siloed systems. The offshore wage arbitrage made it cheaper to retain this model than fixing the tech stack at scale, until AI came about.

The Double-Edged Sword of AI

AI has now burst onto the scene spawning all sorts of disruptions that will re-invent, or in some cases, demise parts or whole industries. Banking operations in India are projected to see productivity gains of up to 46% from Gen AI by 2030, with cost per transaction falling to as low as one-tenth of traditional manual processing in some use cases.

This automatically instills fear in the workforce about the risk that this transition may have on their jobs. If anything, history consistently proves that "godlike" technological shifts that emerged in every industrial revolution (pandemic included) act as a catalyst for net job creation, not destruction. The introduction of ATM machines in the 1970s did not eliminate bank tellers; instead, it slashed the cost of running a branch by 10x, leading to ten times more branches and an eventual increase in the total number of tellers. During the Agricultural Revolution, we did not just replace farmers with machines and leave people with nothing to do; we transitioned from humans doing the manual labor to humans servicing the machines that did 100 times more work.

While the AI infrastructure is moving at pace, we are yet to define how we transition talent to that future AI-enabled workforce and this is where we as GCC leaders should help navigate and facilitate mass reskilling along with AI-native re-engineering vs. earlier human centric process engineering.

Future Workforce Trends

The future workforce is undergoing a "Gutenberg moment", transitioning from the manual "labor" of information processing to the architectural "guidance" of AI-enabled outcomes. I like the way Salim Ismail describes this shift as a “move from scarcity to abundance” where the value of a human worker is no longer found in routine execution, but in complex problem-solving and emotional intelligence. A successful future workforce will be one with the following trends: -

• The 80 / 20 manual processing rule: AI handles 80% of “grunt work” tasks that workers today are required to do in order to achieve a customer outcome, which would then allow professionals to apply human judgment to the "tricky 20%" of scenarios that require high-level exception handling, nuanced decision-making and complex problem-solving.

  
  

• From Volume to Value-based roles: Rather than shrinking the workforce, increased efficiency allows for radical expansion. Just as the ATM machine reduced branch costs by 10x—leading to ten times more branches and more tellers overall—AI-native re-engineering will allow operational staff to move into high-value, customer-facing roles. The focus shifts from the "back office" to the "front line," where staff can leverage design thinking to enhance the customer experience.

  
  

• The 10x capability leap: Because an experienced professional is now 10x more capable than they were just two years ago, they can guide AI to handle complex processing that once required senior oversight. This "juniorizes" technical execution, but it demands a more versatile worker. Staff must possess deep product knowledge and a holistic understanding of a customer's business to effectively "pitch" and curate AI-generated solutions. The goal is for humans to become "caretakers of the AI", ensuring the technology aligns with human-centric goals,

  
  

• The Emergence of AI-Native Specializations: New roles will emerge, centered around AI model trainers, data scientists, platform integrators, compliance as a service, customer experience specialists, AI decisioning middleware – all of which require different skillsets.

All this then begs the question for GCC leaders: how to deliberately transition from high-volume, junior-staff-heavy operations to a smaller, more senior, AI-enabled workforce that owns outcomes, not tasks.

Blueprint for reinvention

For an AI-first GCC, we as leaders must set bold, forward-looking objectives to lead the AI-first transition, turning offshore hubs from execution engines into strategic, innovation-led partners – the shift from “Volume factory” to “Value engine”. I believe the top 5 initiatives should include -

1. Build AI-native talent pipelines: Drive internal academies and partnerships to upskill 50–70% of the workforce into AI literacy, domain-AI hybrids, and model stewardship roles, positioning the GCC as a talent multiplier rather than a consumer. Make this a formal part of objective setting and not a lose checklist exercise to overcome “learning lethargy”. This will also partly address better career planning, improve retention, and drive upskilling momentum.

2. Own end-to-end AI-led journeys: Take responsibility for complete banking processes (KYC, servicing, transactions, payments, risk) using agentic AI, pods, and outcome metrics, proving the GCC can deliver superior speed, quality, and CX at lower cost.

3. Establish robust AI governance and ethics: Create center-wide frameworks for model risk, bias monitoring, human-in-loop controls, and regulatory alignment, building trust with onshore stakeholders and regulators while de-risking scale.

4. Drive innovation labs and IP creation: Launch dedicated AI innovation teams to ideate, prototype, and globalize use cases, shifting from process optimization to creating proprietary tools, digital products, and competitive edges for the enterprise.

5. Redesign for pod-based, agile operating models: Transition to cross-functional, autonomous pods with shared services, agile delivery, and AI-first infrastructure, enabling faster experimentation, continuous learning, and adaptability to evolving tech and business needs.

Call to Action: Lead or Follow

The AI reckoning is here. As GCC leaders, we should not defend the past, but architect the future. Start with one pod, one journey, one academy today.

---

About The Author

Anil Nayak is a senior banking executive with deep experience across client servicing, operations, and risk-led transformation. Currently a Managing Director at HSBC, he leads large, multi-country teams focused on client onboarding, KYC, credit, and data-driven operations. His work sits at the intersection of AI, operating model design, and the future of financial services, with a strong focus on shifting teams from repetitive processing to higher-value, outcome-driven roles.

Disclaimer: Any views expressed in this article represent the author's personal opinions and not of his employer.