How AI is reshaping enterprise systems, global operations, growth, and talent—and what GBS leaders can learn from those putting it to work.
AI is opening up new choices for Global Business Services leaders: where work happens, how teams support a growing business, which systems deserve investment, and how the next generation of talent develops. Those choices could have a far greater impact than the savings from any single automated task.
Yet most organizations are still early in that shift. According to ABSL’s Business Services Industry Global Report 2026, only 2% run AI agents at scale. What can leaders learn from those already putting AI to work?
At the ABSL Forum, Michael van der Steen, who leads Cencora Business Services, and Ana-Maria Constantin, CEO and co-founder of Cascade, explored where AI is delivering value and how it is changing decisions about the way GBS operates.
Their discussion surfaced five lessons:
- AI Creates Value When It Is Matched to the Work. The organizations getting results treat AI as one tool among several, using it where reasoning helps and keeping deterministic methods where exactness is required.
- Intelligence Is Moving to the Front of the Enterprise Stack. As AI agents handle more of the logic of how work gets done, employees get a simpler way to work and enterprises get new choices about how they design and invest in the systems underneath.
- AI Translation Gives GBS Leaders More Optionality in Global Operations. AI translation lets a team working in one language support customers and employees in another, giving GBS leaders far more freedom in how they organize work across their global centers.
- AI-Enabled Services Helps Companies Support Rapid Growth. A support organization that absorbs new people, entities, and policies without adding headcount in proportion lets a company integrate acquisitions and expand faster.
- AI Is No Reason to Stop Hiring Early-Career Talent. Graduates who arrive with drive and the right attitude are thriving in teams that work with AI, which is why graduate hiring and talent programs remain a priority.
The thread running through all five is that the opportunity is moving from the cost of each task toward the design of the work itself.
AI Creates Value When It Is Matched to the Work
Every technology wave arrives with the suggestion that it changes everything. Experienced operators know that a new capability is valuable where it fits and expensive where it does not.
“The way I look at it is AI is another tool in the backpack,” says Michael, “and you need to use it appropriately.” His test is one every finance leader will recognize: “Am I going to use a probabilistic tool to do my accounting? Probably not.” Numbers that close the books need to be deterministic. Interpreting an employee’s question against a policy that varies by country and entity is a different kind of work, and it is exactly where a reasoning system helps.
This is why so many AI programs stall between a promising pilot and real scale. The limit is seldom the technology. It is knowledge of how work actually flows through an organization. “You need to understand the end-to-end and orchestrate that appropriately,” Michael says, “and then use the right tools for the right situation.” That understanding is what GBS organizations have spent decades building, and it has become the scarce input.
Intelligence Is Moving to the Front of the Enterprise Stack
For two decades, enterprise architecture has been organized around systems of record. The logic of work lived inside them, and people learned to work the way the system required. AI loosens that arrangement. When an agent can understand a request, apply the relevant policy, and complete the transaction, the logic moves toward the person doing the asking.
Michael describes where that leads. “If you put the intelligence not in the back end, but you start to put that more in the front-end layer, then the back end just becomes a storage of data.” The investment question changes with it: “Why would you spend 800 million to a billion on an ERP system if you can actually today, with the coding that you can do a lot faster, build your own solutions?” He offers this as a question worth asking and not as a plan, and it will weigh differently for a manufacturer whose ERP runs the physical business than for a services company.
Employees are pulling in the same direction. People entering the workforce expect to state what they need and have the system ask the right questions and finish the job. “The mindset is changing,” Michael says. “It needs to be easier. It needs to be simpler.” The return is productivity, because an intuitive interaction is faster to learn than a catalog of transactions.
Two implications follow for GBS leaders. The first concerns budgets. AI agents draw on labor spend as well as software spend, which makes the familiar argument between headcount and ERP cost a single design question. The second concerns durability. Systems of record will change, and so will the way they are priced. The experience of the people who depend on them should not. “You may change from Workday into SuccessFactors. You may change from ServiceNow into Jira,” says Ana-Maria. “The employee experience should be the same, should be consistent.”
An organization that separates the experience of work from the systems that store it keeps its freedom to choose. As nearly every vendor adds AI of its own, someone also has to decide how the pieces fit together without multiplying cost, risk, and confusion. GBS, which already sits across functions, is the natural owner of that decision.
AI Translation Gives GBS Leaders More Optionality in Global Operations
Location strategy has always been an exercise in constraints: talent, cost, time zone, and language. Language has been among the most stubborn. Serving a French-speaking customer meant staffing French speakers, which narrowed the choice of sites and raised the cost of the service.
AI that translates in both directions removes that constraint. “For us, the biggest single change of using AI actually is the fact that we can now use language neutralization,” says Michael. Work that once had to sit in a higher-cost location due to language requirements can now be delivered from centers at roughly 40% lower cost.
The lesson reaches well beyond translation. Michael contrasts this kind of gain with the more familiar hunt for AI value, where a team tries “to find an individual use case where I can cut four FTEs out of 50 in accounting.” A use case trims a process but a change in what a location can do reshapes the operating model.
Leaders who ask only which tasks AI can take over will find the smaller prize. The larger one goes to those who ask which long-standing constraints on their operating model no longer hold.
AI-Enabled Services Helps Companies Support Rapid Growth
For companies navigating rapid inorganic growth, cost and complexity can multiply quickly. New people, new entities, and new policies arrive together, and the capacity of the functions that support employees becomes a quiet limit on the strategy itself. A services organization that can absorb complexity quickly without adding headcount in proportion raises that limit.
OneOncology, a national network of oncology practices, shows what this looks like. The company was expanding by bringing new practices into its network, each with local policies layered on top of the parent company’s rules, while its HR team supported physicians, nurses, and administrative staff from shared inboxes. Working with Cascade, it started with benefits, deploying AI agents that were personalized to each employee, confidential, and always available.
As Ana-Maria states, “Earlier generations of automation typically delivered rates of 30% to 60%, but, with AI that understands the enterprise context, we’re accomplishing north of 90%, and in the case of OneOncology, north of 98% reduction in manual HR transactions.” The work expanded from benefits to HR more broadly and then to IT shared services, as the organization grew more than threefold in two years with only a minimal increase in HR headcount.
OneOncology is now part of Cencora, and Michael sees the effect from the acquirer’s side. “It made a huge difference for us,” he says. “We’re a company that’s acquiring a lot of businesses, so integrating them into the business is fairly critical for us.”
The same logic should govern where GBS leaders choose to innovate. Michael’s advice is to “develop an ecosystem of partners that helps you succeed and fix the problems that you have,” and to judge each one by whether it advances the enterprise strategy. Ana-Maria applies the same standard: “You’re either driving business value or you’re not.”
AI Is No Reason to Stop Hiring Early-Career Talent
While the prevailing narrative suggests that AI’s ability to automate junior-level tasks reduces the need to hire entry-level talent, acting on that assumption risks backfiring in several key ways. It empties the pipeline of future leaders, and it removes the people most likely to redesign the work, because they never learned to accept the old way of doing it.
Michael’s advice to GBS leaders starts here. “Be open-minded. Still hire the young talent. And develop a talent program.” In India, Cencora’s graduate program accounts for about 100 people in a team of roughly 900, “and that really helps us to succeed.”
Ana-Maria sees the same thing inside an AI-native company. Graduates “who have the drive, want to prove themselves, who have the right attitude” are, in her words, “some of the most successful people out there.”
She adds an observation that asks more of leaders. “I’m seeing AI as an enhancer of culture. You give AI in the hands of top performers, the sky’s the limit.” Where standards are weaker, the same tools tend to shift load onto everyone else. AI raises the return on having a strong team, which places hiring, development, and culture among the most consequential AI decisions a GBS leader will make.
The Opportunity Belongs to Leaders Who See the Work End to End
What connects these five shifts is a drive toward better outcomes more than legacy tools, process or playbooks. Technology is now changing faster than most implementation cycles, and attachment to any one system becomes a liability. Ana-Maria’s advice to technology leaders, offered as a fellow engineer, is to “try to not be emotional about your existing software projects” and to stay “more focused on outcomes and results rather than on a specific piece of software.”
None of this arrives automatically. Every organization starts from a different level of maturity, and each will find its own version of these opportunities. The direction, though, is consistent. The value is moving from the individual task to the design of the work, and design rewards the people who understand how work moves across functions, systems, and borders. That has been the particular expertise of GBS for as long as the profession has existed. AI gives that expertise more to work with.
At Cascade, we build AI agents for HR and IT Operations and Global Business Services that work across a company’s existing systems of record, so employees get consistent, confidential support while the organization behind them grows and changes.
This piece draws on a conversation between Ana-Maria Constantin, CEO and co-founder of Cascade, and Michael van der Steen, CEO and Vice President of Cencora Business Services, at the ABSL Forum 2026 in Budapest.



