Build-Operate-Transfer vs. AI-Native GCC Design: Choosing an Operating Model in 2026

Every enterprise standing up or resetting a Global Capability Center in 2026 runs through the same governance decision: build a captive entity from scratch, outsource it entirely, or use a Build-Operate-Transfer model that phases ownership over time. It is a real decision with real consequences for capital exposure, compliance risk, and how fast the center can hire.

It is also not the decision that determines whether the center creates value.

Build-Operate-Transfer is a sequencing model. A partner builds the legal entity, infrastructure, and initial team, operates the center under agreed governance and SLAs, and then transfers full ownership and control to the enterprise once the center hits defined maturity milestones. It exists to solve a specific set of problems: speed to market without the enterprise absorbing early-stage compliance and employment risk, capital efficiency during the setup phase, and a defined path to full control once the center has proven itself.

Those are legitimate, board-level concerns, and BOT is a reasonable answer to them. It says nothing, however, about how the center actually works once it is running.

A BOT contract can transfer a center that runs on agentic workflows and a governed AI platform. It can just as easily transfer a center that runs the same manual, ticket-based processes a traditional shared-services operation ran a decade ago, just with a different ownership timeline attached. The BOT structure is silent on which one the enterprise ends up with.

This matters because the data on what actually drives value points away from ownership structure and toward workflow design. McKinsey’s State of Organizations 2026 research found that traditional centralization models, the “lift and shift” approach that BOT engagements often default to, bring near-term labor cost savings that plateau quickly once the initial migration is complete. Models that automate and redesign the process first, then centralize, deliver more durable efficiency and quality gains. The order of operations matters more than who owns the entity at any given point in that order.

The same research shows enterprises are already running into this gap regardless of governance structure. Eighty-four percent of leaders plan to expand the scope of their shared-services and capability centers in the next one to two years, but more than 40 percent have not started systematically adopting the technologies that expansion depends on, and only 6 percent of business services leaders report realizing full value from advanced technologies across multiple use cases. Ownership structure was not the bottleneck for any of them. Workflow and platform design was.

The real 2026 decision is not BOT versus captive versus managed. It is two independent decisions that most transformation programs collapse into one:

Ownership structure, meaning who controls the legal entity, employment, and governance, and over what timeline. BOT, captive build, and managed models are three answers to this question.

Operating design, meaning whether the center is architected around AI-native workflows, agentic orchestration, and outcome-based governance from the start, or whether AI gets layered onto a center built around traditional process and headcount management.

These two decisions are orthogonal. An enterprise can choose BOT and build an AI-native center from day one. It can also choose BOT and inherit, at transfer, a fully staffed traditional operation that now has to be re-architected after the fact, which is a materially harder and slower change than designing it in from the start. The ownership model does not determine the outcome. What gets specified in the build phase does.

If BOT is the right ownership structure for an enterprise’s speed and risk profile, the AI-native architecture needs to be a deliverable of the build phase, not a post-transfer initiative.

McKinsey’s research offers a concrete example of what that looks like in practice: one large bank used hybrid “digital factories” for legacy application modernization, in which human workers were elevated to supervisory roles overseeing squads of AI agents rather than executing the modernization work manually. That is a build-phase decision. It has to be architected into the center before operations start, because retrofitting agentic orchestration onto an already-staffed, process-heavy center is a much larger undertaking than designing it in from the outset.

The same logic applies to governance. If the operate phase is managed against headcount ratios, SLA adherence, and cost per FTE, that is the scorecard the enterprise inherits at transfer, along with a center architected to hit those specific numbers rather than business outcomes. Outcome-based governance, cycle time, defect rates, time to market, needs to be the standard from the first day of operation, not something introduced after the enterprise takes control.

The platform layer matters just as much. A governed, reusable AI platform, comparable to Ascendion’s AAVA™ for engineering workflows, needs to be built as transferable enterprise IP during the build and operate phases. If the AI capability lives in a partner’s proprietary tooling that does not transfer with the entity, the enterprise inherits a center that has to rebuild its intelligence layer from scratch the moment it takes ownership.

None of this makes the ownership question irrelevant. Capital appetite, regulatory complexity by geography, and how quickly an enterprise wants full operational control are still real inputs into whether BOT, captive, or managed is the right starting structure. McKinsey’s research also points out that hub-and-spoke models, which balance central efficiency with local flexibility, remain relevant, and that centers increasingly operate as virtual, distributed capability rather than a single physical location tied to one jurisdiction. Location and legal structure still carry real geopolitical and compliance weight, particularly as the same research found that almost three in four leaders report geopolitical uncertainty is already affecting their organization.

The point is not that ownership structure doesn’t matter. It is that ownership structure answers a different question than the one that determines whether the center creates value.