Shot of two programmers working together on a computer code at night

How agentic AI can transform GBS into an enterprise intelligence hub

Authors

23 Jun 2026

GBS is at an inflection point as agentic AI scales. Leaders must act now to support end-to-end, agent-led execution.

In brief
  • GBS organizations are at a structural inflection point as agentic AI moves from experimentation to operational reality, changing how work is done.
  • Leaders must make foundational decisions now across data, capabilities, operating models and technology — regardless of how the AI future unfolds.
  • Unlocking agentic AI requires a structured, outcome-driven, layered approach to reimagine processes and support agent-led execution at scale.

Over the past two decades, global business services (GBS) has been a key driver of cost efficiency, built on standardization, automation and labor arbitrage. According to the EY-sponsored 2023 Shared Services & Outsourcing Network (SSON) survey, leading GBS organizations have already moved beyond this model, positioning themselves as platforms for broader enterprise value creation with 81% actively supporting enterprise digital agendas.

However today, many GBS organizations stand at a structural inflection point. The question is no longer how to drive incremental efficiency improvements, but how GBS organizations fundamentally reimagine how work is executed end to end with outcome and customer experience at the center.

The development with agentic AI is only reinforcing this inflection point. It moves GBS beyond task automation toward intelligent orchestration, where AI systems can reason, make decisions, and act across processes and systems. But the opportunity is not unlocked by technology alone. It requires new operating models, clearer process ownership, stronger data foundations, and different workforce skills and global governance.

From AI experiments to real enterprise value

Despite unprecedented attention and investment, many organizations have yet to realize the full value of AI. When we started working with AI solutions based on large language models in early 2024, it made sense that most initiatives were proof of concept (PoC). The focus was on testing the technology, learning fast and building experience. Experimentation was the right place to start.

Two years later, however, we still see many AI initiatives stuck in PoC mode, never moving into production. Even when AI is deployed, it is often limited to chatbots that provide information to end users. These solutions can deliver some efficiency gains, but they are typically incremental. They depend on users actively choosing to engage with the AI instead of working the way they always have, which limits adoption and impact. Many AI solutions are still point solutions, bolted onto legacy processes not designed for AI, while organizations consistently underestimate what it takes to make AI work at scale — from data and infrastructure to governance and iteration with the business.

To unlock real value, we need to change our starting point. Instead of searching for isolated AI use cases, we should look at end‑to‑end processes and ask how they should work if AI is at the core. That means reimagining processes rather than enhancing them, designing explicitly for zero‑touch or less human intervention where it makes sense, and then intentionally bringing humans back in where they add the most value. This shift — from bolt‑on solutions to AI‑driven process redesign — is what will move organizations from experimentation to transformation.

Leading through uncertainty while building the foundation

One thing is certain about the AI journey; no one knows exactly how the future will unfold and as a result organizations, leaders, and employees today perceive the impact of AI in very different ways. The EY Four Futures of AI illustrates exactly this — meaning multiple possible scenarios for how AI could reshape the business landscape by 2030. While these futures differ in pace and form, they share a common theme, AI becomes deeply embedded into core business processes.

Leading in this level of uncertainty is challenging and organizations must move forward without having full clarity on the destination to keep competitive. Yet, despite this uncertainty, there are some foundational decisions leaders must make today to support their AI journey:

  1. Data foundation — leverage data as key competitive advantage
    A strong, secure and accessible data foundation is not only critical, but also a true source of competitive advantage. Investing in clear data strategy and platforms that support secure, well‑governed access to high‑quality data will be essential.

  2. Culture and capabilities — help and prepare people for the future
    Successful transformation is only possible when people are fully onboard and equipped with the right capabilities. Investing in skills, continuous learning and change readiness is just as critical as investing in technology.

  3. Operating model — built for an intelligent enterprise
    As technology advances, organizations need to support operating models to evolve to drive greater value, agility and innovation. By anchoring design decisions in business outcomes, organizations can shape processes and workflows around value rather than technology. This entails reimagined processes and workflows through horizontal and vertical agents to capture value end to end, as well as continuous evolution of roles and capabilities as work shifts from task‑based AI to agentic operations.

  4. Technology foundation — create a flexible technology ecosystem
    A flexible partner ecosystem, combined with a modular and lean technology platform, preserves optionality and helps integration of new AI capabilities as the landscape evolves. This adaptability is essential for scaling agent‑based solutions over time.

Regardless of which AI future materializes, the implications for leaders are clear, decisions made today matter and getting on these foundations are essential to remain competitive in the future. 

How EY can help

Agentic AI is no longer a future bet

For a long time, AI capabilities were promising but fragmented. Solutions were not enterprise ready, hard to govern, and difficult to scale. Recent advances in agentic AI mark a clear shift from experimentation to operational reality.

 

What has fundamentally changed is the ability to design specific AI agents that are very good at one specific capability and stable enough to be embedded directly into core processes. At the same time, market standards for communication and integration are beginning to stabilize. Common ways for agents to interact with enterprise systems, external tools and other agents are emerging, making it possible to automate larger parts of end‑to‑end processes without unnecessary human interaction.

 

This is no longer a five‑year horizon: waiting risks locking organizations into architectures and operating models that are not designed for agent‑based execution, creating both technology and competitive debt.

 

From shared services to an enterprise intelligence hub

As agentic AI moves from experimentation to operational reality, organizations must decide where and how this capability should be anchored to deliver value at scale — and GBS is positioned to take on that role.

 

With established end‑to‑end enterprise process ownership, GBS can imagine how work is executed across functions, rather than improving tasks in isolation. Its central role in supporting high‑quality, enterprise‑wide data and system access, combined with mature governance and controls, provides the trusted foundation agentic AI requires to reason, decide and act across the organization.

 

GBS also brings deep improvement and automation experience, built through years of standardizing, orchestrating and scaling complex operations. This helps reusability by design, allowing shared AI capabilities, agents and orchestration patterns to be developed once and reused many times, compounding value over time.

 

Together, these strengths position GBS to evolve beyond a service delivery model into an enterprise intelligence hub, orchestrating people, processes and AI agents at scale.

 

From vision to value through a 7-layer framework for scaling agentic AI

To move from isolated AI initiatives to real enterprise value, organizations need more than technology. At the EY organization, we use a structured framework designed to move organizations from bolt‑on AI solutions to AI that is built into core processes: EY.ai Value Blueprints.

 

The framework starts with end‑to‑end processes, not use cases. We define the objective and the jobs to be done within a process. Each job is then mapped to the AI skills required to perform it, such as reasoning, decision support or execution, and assessed for AI readiness based on factors like volume, risk and data dependency.

 

By identifying overlap in AI skills across jobs to be done, we can redesign processes end to end rather than enhance them step by step. The redesign follows a design‑for‑zero approach: Reduce manual interaction by default and deliberately place humans in the loop where risk is high, accountability is required or human judgment provides a competitive advantage.

 

The reimagined processes are implemented through seven interconnected layers, spanning customer experience, human‑AI collaboration, AI‑first processes, trust and responsible AI, enterprise intelligence, AI‑native platforms, and systems of record. Working together, these layers support scalable, governed AI and allow value to compound across processes over time.

Summary 

Agentic AI is moving from possibility to operational reality, fundamentally changing how enterprises work. Now is the time to lay the right foundations, across data, capabilities, operating models and technology, while reimagining how work is executed end to end to fully realize the value of agentic AI and support GBS to lead as an enterprise intelligence hub. The question is not whether AI will change GBS, but who will act now to shape the change.

About this article

Authors

Related topics
AI
Consulting
Finance

EY refers to the global organization, and may refer to one or more, of the member firms of Ernst & Young Global Limited, each of which is a separate legal entity. Ernst & Young Global Limited, a UK company limited by guarantee, does not provide services to clients.