Download the report
Our take
GBS and agentic AI: Friends or foes?
From redefining performance metrics that measure business outcomes to building robust governance frameworks that uphold AI accountability, agentic AI is reshaping the core architecture of global business services (GBS), especially within global capability centres (GCCs). Tanuj Talwar and Vivek Chandra provide the lowdown.
Agentic AI in GBS
The shift in the value equation
An integral part of GCCs, GBS is undergoing a step change. By harnessing autonomous technologies, GBS leaders are scaling transformation across their global operations, generating measurable outcomes.
Automated procurement platforms are streamlining supplier onboarding and verification processes, while enabling AI‑driven contract management and compliance monitoring. Predictive analytics is optimising spend and mitigating supplier risk across these centres. In finance, autonomous systems are slashing process cycle times by up to 70% across diverse functions. Meanwhile, back‑office operations are being re‑engineered through intelligent process automation (IPA), reducing errors and driving significant productivity gains.1
These examples signal a clear pivot away from heavily human-dependent processes in favour of AI-powered autonomous agents working in tandem with people to achieve greater speed and precision.
Such value shifts have encouraged organisations enabled by GBS to reassess their agentic AI investment strategies. Research suggests that these organisations stand to gain greater competitive differentiation by building comprehensive agentic AI applications and ecosystems, rather than limiting their efforts to isolated task-specific experiments.2
However, more autonomy brings more complexity and risks and therefore demands more accountability from the GBS unit. Organisations need to optimise their investments and value generated by agentic AI integration with strong underpinnings of responsible AI, ethics, governance, and holistic assessments of human–AI teams. This report examines how agentic AI is driving the evolution of GBS—highlighting the need for AI-ready governance models and redefining talent and leadership for a technology-driven future.
Besides cost efficiencies, one of the primary considerations is value creation for business impact. The integration of these intelligent autonomous AI systems in core workflows unlocks new frameworks of working, revenue creation, and scalable service delivery.
Agentic AI: A game changer for GBS
Recent studies by PwC India have found that global capability centres (GCCs) in India have achieved a relatively high level of maturity in operating as cost-conscious innovation hubs and multifunctional centres of excellence for their headquarters (HQ). Our research indicates that many business leaders in the HQ are upbeat about their Indian GCCs, with GBS serving as the cornerstone, increasingly evolving into go-to hubs for core technology and business processes, driven by the accelerated adoption of AI and digital technologies.3
In fact, in the last five years, GCCs have increasingly leveraged hyper automation and AI-driven automation at scale. Moreover, amidst ongoing economic uncertainty—particularly around tariffs and visa regulations—US firms are reportedly accelerating the shift of critical operations to India, fuelling the rapid expansion of GCCs.4 This trend underscores the importance of not merely expanding the headcount in GCCs but adopting agentic AI especially within their GBS to enable smarter, scalable, and more resilient operations through intelligent automation.
In line with this progression, the GBS organisation too has evolved, embedding ideation and innovation to significantly move up the value chain. It now delivers end-to-end services and exercises greater autonomy in decision-making across technology, architecture, finance, and operations. The new GBS is now being integrated into the HQ, leading to a transition of HQ roles into the GBS structure. This shift is largely driven by the rise of AI agents, which enable GBS units to handle complex tasks across enterprise platforms and transform the economics, speed, precision, and resilience of operations. As multiple AI agents operate independently or collaboratively to coordinate and execute processes across diverse ecosystems with minimal or no human intervention,5 they accelerate innovation and enable hyper-personalised services at scale.
Maturity level of autonomy in agentic AI
Level 1-2
Assistive/orchestrated – agent drafts, human executes
Level 3
Semi-autonomous – executes within guardrails; human approves exceptions
Level 4-5
Autonomous/swarm – multi-agent collaboration; decision logs + guardrails essential
Level 1–2
Assistive/orchestrated – agent drafts, human executes
Level 3
Semi-autonomous – executes within guardrails; human approves exceptions
Level 4–5
Autonomous/swarm – multi-agent collaboration with decision logs and guardrails
Agentic AI’s pivot beyond ‘assistive’ Gen AI to autonomous goal-driven execution is helping reshape GBS from a transaction engine to a business-impact partner. In the process, it is enabling GBS’s move up the maturity curve as go-to hubs for core technology and business processes.6 A case in point is a multinational human capital management (HCM) cloud company that has taken the lead in leveraging agentic AI to scale down heavy reliance on people and spreadsheets. Cutting down turnaround times to a matter of minutes, this move is helping reshape key HR processes such as complex payroll-related tasks. In another instance, a leading retailer has turned its focus on AI agents that can carry out complex industry-specific tasks involving software in a comprehensive manner.
Our take
Disruptions impacting GBS
Levers of change: From shared services to cognitive enterprises
In today’s digitally augmented business environment, technology is no longer limited to supporting operations—it is actively shaping how decisions are made, executed, and scaled. Intelligent platforms are increasingly taking on responsibilities that go beyond automation, acting as collaborative partners to human teams across service functions.7
For instance, a banking technology platform provider brought in agentic AI to offer proactive financial management support to consumers in a bid to elevate their overall experience. The differentiator is the AI agent’s capabilities to serve consumer and banking needs in the areas of fraud detection, marketing, risk management, lending decisions, and back-/mid-office operations.
As organisational maturity increases, the value contribution of GBS is shifting decisively. The earlier model, anchored in workforce scale and cost efficiency, is being replaced by one that prioritises measurable business impact such as revenue enhancement, spend discipline, and risk resilience.
The deeper integration of digital agents into enterprise operations allows GBS organisations to move from execution support to leadership. These AI-led capabilities enable:
Accelerated innovation turnarounds
Decision support powered by predictive insights
Personalised experiences delivered at enterprise scale with lower dependency on manual intervention
For organisations preparing for this shift, GBS is increasingly transforming into a strategic engine for resilience, agility, and value creation, rather than a traditional shared services construct.
Redefining the GBS operating model
While agentic AI delivers 24×7, is scalable, proactive, autonomous, capable of real-time decisions, and can delegate more decision-making to technology, with new technology new problems arise. These new challenges include:
- Introducing and maintaining agents in an evolving business and technology ecosystem
- Aligning agents with business priorities
- Ensuring compliance with regulations
- Maintaining transparency and traceability to avoid ‘black box’ scenarios
- Eliminating hallucinations to maintain end-to-end accountability
Amplifying the value of GBS
As agentic AI rapidly becomes embedded within GBS, a clear shift from traditional cost-arbitrage and volume-based cost models towards outcome-driven approaches is palpable. The following are key transitions that can unlock significant value:
- From cost efficiency to value creation
- From activity metrics to outcome indicators
- From decentralised controls to unified AI governance
- From functional roles to techno-functional leadership
- From experimentation to leadership commitment
From cost efficiency to value creation
GBS is no longer viewed primarily as a cost lever. It is evolving into a strategic innovation layer that contributes directly to revenue generation, operational risk reduction, and enterprise resilience.
From activity metrics to outcome indicators
Traditional SLAs and throughput metrics are giving way to performance measures that track knowledge outcomes, business value, and experience improvements across functions.
From decentralised controls to unified AI governance
As AI autonomy increases, organisations must establish centralised governance structures that enable visibility, auditability, and ethical oversight across all AI agents and decision layers.
From functional roles to techno-functional leadership
The workforce supporting agentic AI needs to blend domain knowledge with AI fluency. Roles such as AI product owners, agent engineers, and responsible AI leads become critical in governing human-AI collaboration.
From experimentation to leadership commitment
Scaling agentic AI requires more than pilots. It demands executive sponsorship that aligns funding, governance, and incentives directly with business priorities.
By embedding autonomous agents into core workflows, organisations will not only streamline execution, but also build the foundation for scalable, intelligent operations that drive measurable business outcomes. Our recent research found that 66% of respondents have seen productivity gains and 57% saved costs. Also, 55% accelerated decision-making and 54% improved customer experience from agentic AI implementations.8
Agentic AI is therefore gradually transforming the delivery model by automating high-volume, repetitive processes end-to-end. This shift enables faster cycle times, reduces SLA breaches, and enhances operational resilience—ultimately supporting ‘always-on’ service delivery. API-based billing, premium pricing on AI products, and bundling of multiple offerings are other examples of business models that are coming into vogue.
Talent evolution in the age of agentic AI
With work being redefined, the traditional functional specialisation within GBS too is making way for a new paradigm that blends deep domain expertise with AI fluency, data literacy, and governance capabilities. In this environment of ‘techno-functional’ talent, the role of IT services professionals will evolve from executing tasks to engaging in roles focused on supervising agents, aligning policies, mitigating risks, and developing hybrid skill sets that factor in:
| AI product ownership | Human-in-the-loop (HITL) design | AgentOps |
| Bias mitigation | Control monitoring | Deep domain knowledge in areas such as finance, HR, and CX |
These techno-functional profiles could be instrumental in ensuring AI agents operate within defined policy boundaries, generate measurable outcomes, and augment rather than replace human capabilities.
To accelerate this transformation, organisations must adopt a structured talent strategy that includes:
GBS organisations must invest in capability-building initiatives that teach AI tools, prompt engineering, analytics, and risk governance, particularly to domain experts. Rolling out programmes that transition operations analysts into roles such as AI engineers or product owners can enhance talent retention.
Conventional roles are evolving into newer archetypes such as functional AI product owner and process and agent engineer. These roles must be integrated into workforce planning and compensation frameworks.
Human-agent collaboration will only succeed if employees understand, trust, and effectively work alongside AI. Change programmes must therefore focus not just on skill but also on mindset.
As hybrid human–AI teams become the norm, HR functions must evolve to manage their performance, workflows, and governance. This includes defining agent role descriptions, assessing AI outcomes, and integrating digital labour into organisational design.
The success of agentic AI in GBS hinges on how well organisations can build a digitally confident, outcome-focused workforce. Metrics such as time-to-proficiency, AI role coverage across value streams, policy adherence, and measurable KPO gains will define the maturity of techno-functional talent ecosystems. In this agentic future, GBS talent will no longer be defined by cost effectiveness, but by its ability to unlock business value through intelligent orchestration of humans and AI.
Translating thought into action
As agentic AI transitions from conceptual pilots to production-scale deployment, GBS will need to shift focus from transaction throughput to delivering business outcomes. The leap is not simply about automating tasks. It means embedding autonomous execution with governance, outcome-based KPIs, and techno-functional leadership to drive growth, improve cash flow, elevate CX, and build trust.
Prioritising value streams aligned with business priorities, clear ownership and traceability across processes, implementing runtime guardrails, and deploying pilot AI agents with well-defined checkpoints are critical to mitigate model, operational, and regulatory risks. It implies the need for a resilient and responsible strategy encompassing:
AI product owner who owns the agent design, decision logs, model cards, and release gates
Business owner who owns outcomes and policy intent in AI governance
Risk and compliance officer who monitors compliance and enforces standards with guardian agents
Internal auditor who validates traceability and incident readiness in AI systems.
Figure 1: Game plan implementation
Source: PwC analysis
By executing this game plan within a safe, structured framework, GBS can unlock tangible value across different dimensions. The following illustrative examples show how agentic AI can drive measurable impact ranging from improved customer satisfaction and cash flow to enhanced operational efficiency and trust.
Growth and CX
Higher customer satisfaction score (CSAT), improved upsell, greater containment
Cash and spend
Lower days sales outstanding (DSO), improved working capital, higher compliance
Efficiency
Reduced cycle time, increased straight through processing (STP), fewer error rates
Trust
≥99% of actions logged, reduced policy violations, improved transparency
Over time, the new GBS is likely to transform into an autonomous, outcome-oriented function capable of making critical business decisions independently. It will increasingly minimise dependence on human-driven processes while delivering innovation and resilience at scale. However, the true benchmark of success will be how effectively GBS leverages AI to drive growth, enhance efficiency, build trust, and ensure long-term business resilience.
Strengthening governance frameworks
Establishing a traceable AI framework is critical. This involves designing agents or agentic platforms with explainability at their core, supported by robust governance structures that ensure transparency and accountability. Agentic AI fundamentally redefines the questions of ‘who did what, when, and why’. Without clearly defined ownership and risk mitigation measures, traceability mechanisms, and runtime guardrails, initiatives risk stalling at the proof-of-concept (POC) stage.
Moreover, isolated experiments with AI agents will fall short of delivering meaningful competitive advantage. True differentiation lies in adopting a comprehensive agentic AI strategy that reimagines entire ways of working. However, broad adoption alone does not guarantee deep impact. While embedding agentic AI capabilities into enterprise applications can yield localised productivity gains, it may not translate into transformative outcomes without a strategic, enterprise-wide approach.
Our AI Agent Survey 2025 of 300-plus senior executives in the US reflect this view—nearly three-quarters (73%) asserted that it is their manner of using AI agents that will gain them a significant competitive advantage over the next 12 months. An estimated 46% were worried that their organisation was likely lagging behind the competition in agentic AI adoption.9
Research suggests that widespread transformation is possible when entire operating models are created that deeply embed and orchestrate multiple AI agents which collaborate across internal networks and the larger partner ecosystem.
Yet, at present, only a few early movers have shown readiness to make sweeping changes to this degree.
There are some clear deterrents. With greater autonomy comes greater risks. Industry experts believe since agentic AI systems influence real-time decisions, any unintended outcomes can prove damaging. In fact, concerns around compliance may hinder adoption of the technology in highly regulated or less-digitised sectors.10
Our global survey on AI agents revealed that respondents trust agentic AI to handle tasks such as data analysis (38%), performance improvement (35%), and daily collaboration with human team members (31%), but this trust declines to only 20% when financial transactions or other such higher-stakes functions are considered. Concern over cybersecurity emerged as the biggest deterrent in realising value from agentic AI.11
Fostering trust and transparency
As organisations move deeper into autonomous AI adoption, governance will increasingly become a board-level accountability area rather than a technical back-office function. The value of agentic AI lies not only in efficiency and speed, but also in the organisation’s ability to maintain trust, human oversight, and ethical alignment.
Traditional AI governance models must evolve to handle systems where machines not only suggest actions but execute decisions. This requires treating AI agents as enterprise assets that demand the same level of monitoring, control and risk management as core digital infrastructure.
Modernising organisational frameworks, roles, and core processes must prioritise transparent communication, inclusive stakeholder participation, and iterative design thinking principles.12 These elements are essential for effectively navigating the change dynamics that accompanies agentic AI implementation. From the outset, organisations must proactively embed trust and risk management by design into their AI strategies—designing for responsibility and resilience to maximise business impact.13
Governing the lifecycle of AI agents necessitates a cross-functional approach involving automation experts, cybersecurity specialists, legal advisors, risk and compliance teams, and domain experts. These stakeholders must define the rules of engagement—how AI agents are on-boarded, managed, assessed, upgraded, and retired.
Leadership frameworks must also evolve to foster collaboration and orchestration in the era of agentic AI. Leaders will no longer remain simply decision-makers—they must act as enablers and integrators, skilfully balancing human and AI contributions to create synergistic value in human–AI value propositions.
A well-defined governance model can serve as the foundation for future-ready digital teams that act as workforce multipliers, seamlessly collaborating with human counterparts.14 Organisations that master the equilibrium between human oversight and AI autonomy will be best positioned to convert agent-driven insights into strategic decisions at the boardroom level.15
Shaping tomorrow’s GBS
As agentic AI reshapes the future of GBS, its impact calls for a collaborative alignment between global GCCs, market forces, and GBS teams to reimagine operating models, leadership structures, and execution strategies for a future-ready enterprise. As new roles emerge, traditional functions will be redefined—for example, in finance, the role of an accounts receivable lead may evolve into that of a chief revenue manager. Ultimately, every function will require dedicated representation within the re-architected GBS, aligning with the capabilities and demands of agentic AI.
Figure 2: Role archetypes for an AI-first GBS
Functional AI product owner
Defines AI use cases, translates business intent into agent workflows, and owns value delivery.
Data product and analytics lead
Ensures reliable data, performance insights, and outcome visibility for AI-driven services.
Process and agent engineer
Designs AI-enabled processes and builds agents with embedded human oversight.
Knowledge engineer/prompt designer
Structures business knowledge into clear instructions for consistent agent behaviour.
Responsible AI and risk lead
Oversees ethical AI use, risk controls, and compliance across agent deployments.
AgentOps/digital labour site reliability engineering (SRE)
Runs and monitors live agents, ensuring stability, performance, and policy adherence.
Source: PwC analysis
The traditional ‘labour pyramid’ model, built on layers of transactional work, will also be steadily replaced by an AI-first delivery spine. In this emerging model, autonomous agents will execute routine processes at scale, while human teams focus on strategic oversight, exception handling, and innovation. To foster transparency and trust, centralised AI governance must be embedded, reinforced by decision logs and audit trails.
Equally important is the shift from managed services to managed value. Organisations must re-architect their KPIs, moving away from volume- and service-based metrics towards function-specific outcomes—for example, prioritising spend optimisation over the sheer number of purchase orders processed.
A critical leadership mandate is to align the business, technology, and risk functions by bringing together functional leaders, technology architects, and compliance experts. This integrated approach ensures that agentic AI is deployed ethically and securely, without compromising speed, scale, or innovation.
In conclusion
It is clear that shaping tomorrow’s GBS requires a fundamental rethink of its evolving business case in the era of agentic AI. Responsible agentification demands rigorous oversight, robust governance, accountability, transparency, and talent transformation. Only then can the evolved GBS transcend its legacy as a cost-centric process engine to emerge as an extended business function powering key business outcomes.
View earlier editions
Global capability centres
Follow PwC India
Contact us
