Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

AI makes silos impossible to ignore

How your org chart fails your AI investment.

[Photo: Getty Images]

Profile photo of Darren Heffernan,

CEO · Trintech

Over the past year, I’ve had a version of the same conversation with CEO after CEO: “We’ve invested in AI. So why doesn’t the business feel more connected?”

It’s a fair question. AI has accelerated workflows, improved productivity, and automated tasks that once required significant manual effort. Yet many leaders expected something more transformative. They expected AI to naturally connect teams, simplify decision-making, and dissolve organizational friction on its own.

In my experience, the answer lies within the organization.

How AI exposes silos

When companies embed AI into everyday workflows, technology surfaces existing inconsistencies. Sales defines success one way, finance defines it another, and operations relies on different data than procurement. Without a shared understanding of the “truth,” AI amplifies those differences.

For example, I’ve seen organizations where the sales team defines “revenue” as future bookings while finance defines it as recognized income. A simple question like, “How much revenue did we generate last quarter?” can produce different answers depending on which system—and which definition of revenue—AI uses.

Organizations compensate for misalignment by using spreadsheets to bridge data gaps or reconciling conflicting numbers in meetings. Manual reviews paper over inconsistent processes. These workarounds keep the business moving yet hide the underlying problem.

AI depends on standardized data, clearly defined ownership, and consistent processes. Where those foundations are missing, the cracks show up almost immediately, especially as companies move from isolated pilots to enterprise-wide adoption.

Automating one workflow is straightforward. Scaling AI across finance, sales, operations, procurement, HR, and customer service is a different problem entirely. Decisions in one function start influencing another in real time, and accountability is shared rather than owned by a single team.

I’ve seen organizations adopt the same AI capabilities with very different outcomes. In one case, leaders aligned on shared definitions for key business concepts like revenue and customer status, clarified ownership of the underlying data, and established governance for resolving conflicting information. AI became a trusted way to accelerate decision-making across functions.

In another case, each department continued using its own definitions and data sources, resulting in conflicting answers. The technology worked as designed, but teams still spent hours reconciling those differences because there was no agreed-upon source of truth.

Two companies can deploy nearly identical AI capabilities and land in completely different places. The difference is whether leaders aligned their teams around shared definitions and clear ownership before asking AI to connect the business.

Where leadership comes in

AI’s next phase creates a leadership problem, because the organizations getting the most out of AI are the ones that did the unglamorous work first. This includes aligning teams around shared definitions, simplifying processes, clarifying who decides what, and making sure trusted information flows across the enterprise.

They’re also rethinking the relationship between people and technology itself. They treat AI as a coworker that handles the repetitive work and notes what matters, so people have more time to apply judgment where it counts.

Leadership must remove friction identified by technology. Finance is a useful model here, because it’s spent decades balancing automation with governance. Finance teams learned that speed creates value only when it comes with trust and accountability. Producing accurate financial reporting requires consistent numbers from sales, procurement, HR, and operations, using systems never designed to agree with each other.

Finance is where misaligned business shows up first, which is exactly why finance functions have spent decades building the discipline of a single, trusted view of performance. It’s worth borrowing that discipline.

What to ask before the next rollout

So, what does this look like on a workday morning? Before approving another AI initiative, ask a few questions:

  • Who owns this data, end to end?
  • Do sales, finance, and operations mean the same thing when they say “customer,” “revenue,” or “on-time delivery?”
  • When AI surfaces conflicting answers, who has the authority to settle it?

The answers to these leadership questions determine whether AI accelerates progress or just accelerates confusion. Only leadership can eliminate silos.

Darren Heffernan is the CEO of Trintech.