Authored by Niklas Mortensen, EMEA chief design officer at Designit

AI is dominating boardroom agendas, but in too many organizations, the conversation ends there. The gap between ambition and execution is widening. This is not a sector-specific challenge, but some are feeling the disconnect more than others.

With legacy systems, fragmented accountability and complex decision-making chains, the telco sector has become an early test case for what AI transformation actually looks like under pressure. 

Fewer than a quarter of telco executives believe their organization can successfully scale AI. Yet our research suggests this is unlikely to change quickly, given that just 11% of telco leaders have identified organizational redesign as a priority for AI transformation.

That’s a troubling gap for those at the top of businesses. Boardrooms recognize they cannot scale AI, yet few are addressing the organizational structure that is preventing them from doing so.

That disconnect risks locking businesses in perpetual AI pilot mode, delivering impressive demos but negligible returns.

Why telco is an early warning sign 

If AI was simply a technology challenge, then the telco sector would be well positioned to reap the benefits.

There has been significant investment in digital transformation, with global spending reaching $2.5 trillion in 2024, and forecasted to reach $3.9 trillion by 2027. It also generates vast amounts of data, has sophisticated technology estates, and the space is ripe for potential AI applications, from customer service to network management.

Yet confidence in scaling AI remains low.

It makes telco a useful and important early warning sign for every leadership team. If one of the most digitally mature and operationally complex sectors is struggling to move out of the pilot stage, then the limiting factor is unlikely to be a technology problem.

Our research shows a collision between organizational reality and technology. And that should concern boards from other industries.

AI lands in that environment and immediately runs into the same barriers that slowed every previous transformation: siloed data, unclear ownership, and decision-making processes built for a different era.

In many organizations, AI is being introduced into operating models that were never built to support it. New capabilities are layered onto existing structures, but the structures themselves remain unchanged.

The same conditions exist across financial services, healthcare, manufacturing and retail. Any organization balancing legacy systems, distributed operations and complex governance faces a similar challenge.

Telco is showing us, earlier than most, what happens when you try to bolt AI onto an operating model that was never designed to support it. Yet redesigning that model remains the lowest priority on most AI agendas – and that is a strategic error.

Organizational redesign doesn’t mean a restructure to tick a box. It’s a deliberate rethinking of who owns outcomes and decisions, and how data, teams and processes are connected. 

In practice, this means establishing board-level ownership of AI strategy – not delegating it to an innovation function where decisions lose momentum. It also means defining clear accountability for outcomes, not just deployment, and shifting investment logic away from funding individual pilots toward building the connective tissue – the platforms, processes and people – that allow AI to scale across an organization rather than within it.

We’re seeing this first-hand in our work with a major European telecommunications provider. Rather than starting with identifying more AI use cases for the business, our joint approach began by understanding where accountability broke down, where decision-making was being slowed and where AI was struggling to scale. It meant that we could then create an environment where technology, operating models and people using them could evolve together – providing enterprise-wide value.

The solution isn’t to keep investing more in AI. It’s to invest in the conditions that allow it to scale.

Why disconnect is causing AI transformation to fail 

The most common mistake organizations make is treating AI transformation as a procurement exercise. Buy the tools, run the pilots, announce partnerships, and wonder why the returns never materialize.

Pilots are almost always scoped to solve local problems. A customer service bot here, and a network optimization bot there. Each valuable in isolation, but collectively they change nothing about how the business actually operates.

Without a coherent operating model connecting them, these initiatives generate learnings, but not leverage.

What’s revealing about the telco experience is that this isn’t a resourcing problem. Many of these organizations are spending significantly on AI. The issue is that investment flows into discrete projects that fit neatly within existing structures, rather than challenging those structures. The silos that slowed previous transformations are becoming the boundaries within which AI gets deployed. 

Transformation doesn’t work that way. The organizations making genuine progress aren’t running more pilots – they’re asking ‘what changes do we need to make before AI can actually scale?’ 

That shift from “deployment thinking” to “operating model thinking” needs to come from the top.

Why the boardroom needs to think differently

This is where boards need to act differently. AI strategy cannot sit with a Chief Digital Officer or an innovation function alone. It requires board-level ownership, with the same rigor applied to governance and accountability as any other material business risk.

Two things matter most. First, redefining how AI investment is measured – shifting from activity metrics like number of pilots launched and tools deployed, toward outcomes like revenue impact and cost reduction. Second, establishing clear accountability for AI outcomes at executive level, so there is always someone answerable when a program fails to deliver, not just when it succeeds.

However, board-level direction is only part of the challenge. Leadership sets direction and clears structural blockers. But lasting transformation is built around what’s already working at team level. 

The organizations getting this right aren’t just mandating change from the top – they’re building the new operating model on the pilots that have delivered results, the cross-functional ways of working that have emerged organically, and the pockets of genuine AI adoption. These are the proof points that make transformation credible and sustainable. Otherwise, boards risk designing a new operating model that’s just as disconnected from day-to-day reality as the one it replaces.

Telco shows what happens when these things aren’t in place. Investment flows into technology, pilots multiply, and the needle on business performance barely moves. Technology wasn’t holding this back. The operating model was.

Boards that understand this distinction, and act on it, are the ones that will move AI from a promising experiment to a genuine driver of competitive advantage.

The harder question

The telco sector is giving other industries a clear view of what happens when AI ambition outpaces organizational readiness. 

The organizations that come out of this well will be the ones that recognize AI for what it actually is: not a technology problem to be solved by the right tools, but an organizational challenge that needs a different kind of leadership response.

The question for every board isn’t whether to invest in AI. That decision has largely been made. The question is whether the organization is built to make it count.

Photo of Designit’s London office courtesy of Designit