Why telecom firms need a “third space” for AI transformation to really stick

Originally published in SmartBrief

Telecom firms rarely miss a chance to embrace AI.

If there is an opportunity to improve customer experiences, modernize complex networks, reduce costs and tap into new areas of growth, then AI is often top of the list to make these a reality.

Yet the gap between investment in the technology and actual business impact remains stubbornly high. Fewer than one in four telecom executives believe their organizations can scale AI. Designit’s own research, meanwhile, shows that only 11% of telecom leaders think their organizations are prepared for AI transformation.

So while the industry may seem to be moving quickly, this perceived pace should not be mistaken for readiness for change.

Most operators will now have the visible infrastructure of transformation in place. They will have AI strategies, road maps, governance models, technology partners and an ever-growing collection of pilots. All vital.

But these components won’t work without the invisible infrastructure to support them. The shared language across teams, trust between functions, clear decision-making, set responsibilities and the ability to turn promising use cases into a way of operating.

This is where telecom firms need that “third space”.

The gap between strategy and reality

The third space is not another department, innovation lab or layer of governance. It is an environment in which different forms of expertise can navigate ambiguity and surface tensions, and make decisions together before major technical or commercial commitments are made.

The purpose is not collaboration for collaboration’s sake. It is to improve the quality and speed of organizational decision-making, while reducing the risk of scaling the wrong solution.

A business leader may see AI as a growth driver or a way of becoming more efficient. An engineer will focus on performance and the technology’s technical feasibility. The designer? They will be asking how automation affects customers and employees. Meanwhile, operations teams will be concerned with whether it can work consistently at scale.

None of these perspectives is wrong, nor is one more important than the other. But transformation stalls when communication between parties is limited, with these priorities only “vocalized” at the handover stage.

Currently, too many organizations are operating a relay-race model. Strategy defines the ambition; technology builds the systems; design is tasked with making it usable; and the operational organization is finally expected to adopt it. By the time difficulties emerge, each function has already made decisions that impact and constrain the next.

This is not simply a workflow problem. It reflects established habits, incentives and power structures: who is permitted to define the problem, whose expertise carries the most weight and which risks the organization is prepared to acknowledge.

A third space replaces baton handovers with a shared-ownership approach. It gives people permission to expose tensions early, whether it’s a technically impressive solution that creates customer friction or an experience that current systems can’t handle; these will emerge sooner in the process, allowing them to be addressed coherently with a lower chance of fumbles.

People adopt AI, not organizations

This is especially important in telecommunications. Operators are large, highly regulated organizations built on hugely complex legacy infrastructure. Customer journeys cut across channels, platforms and internal teams. Even the smallest AI implementation can affect data, compliance, service design, employees’ roles, and commercial priorities simultaneously.

Looking at this challenge from an anthropological perspective can reveal why these transformation programs succeed or fail. To put it simply, it is not the organizations that adopt AI, but rather the people. Uptake is shaped by whether employees trust the system, how it affects their status within the business, what behaviors are rewarded and whether leaders permit people to challenge the output.

These factors are less visible than the technology, but perhaps play the most important role in determining whether it creates value.

The barriers to adoption are not necessarily new. AI often exposes organizational weaknesses that were already there—fragmented ownership, conflicting incentives, slow decision-making and functions that struggle to navigate ambiguity together. What has changed is the pressure being placed on them. Technology is advancing faster than many organizations can make coordinated decisions about how to use it.

A pilot can demonstrate that an AI tool works. It cannot resolve who owns the outcome, how frontline teams will use it or what happens when it is wrong. These are questions of organizational design and behavior, not simply deployment.

Which is why the rush to implement AI at scale is often counterproductive. Scaling an unresolved pilot will not eliminate its weaknesses; instead, it will spread them across the organization. Operators should not be asking whether a system can be expanded, but whether the organization has developed the shared understanding and operating conditions required for it actually to work.

What a third space looks like in practice

The objective of the third space is not simply to bring different disciplines into the same room, but to create shared ownership of decisions across strategy, technology, design and implementation.

In our work with Dutch telecom giant Odido, where we created such an environment with our parent company Wipro, we agreed on what value looked like and tested decisions against both technical reality and human experience. We also reframed “disagreements” as useful information – encouraging differences not to be smoothed over but instead treated as an opportunity to understand what these different perspectives reveal about the system being changed.

This approach accelerated decision-making, clarified ownership and reduced the risk of problems emerging once the implementation is in full flow.

However, it remains an iterative process.

For leaders, creating a third space is more than just getting a cross-functional team in the same room. People need a shared outcome, genuine authority to make decisions, and continuity to learn together.

AI transformation does not fail only in the technology stack but in the spaces between teams, decisions and incentives. Designing that space may be the most important piece of infrastructure for telecom firms seeking real business transformation.