An article by Rayan Salha, Director Product Marketing at Infovista. 

Artificial Intelligence is fast becoming foundational to the telecom industry’s evolution. However, unlike prior transformations, AI is not just about deploying new systems or upgrading networks, rather it is a wholesale rethinking of how telcos operate, collaborate and make decisions. At FutureNet World 2025, the need for CSPs to become AI-ready was a recurring theme, and one addressed in detail during a fireside chat on-stage between Anita Tadayon, Director of Planning, Transformation and Performance at VMO2, and Rick Hamilton, Infovista’s CEO.

This blog explores the often-underestimated human and structural dimensions of AI adoption in CSPs.  From leadership clarity to cross-functional governance, and mindset shifts to skills development, we’ve drawn on discussions to outline what it truly takes to build a telco that’s ready to thrive in the AI era.

1. AI isn’t just a technology shift, it’s a cultural reset

Unlike the rollout of 5G or even the move to hyperscale cloud, AI doesn’t come with a fixed architecture, a single team in charge or a standard set of best practices. Instead, AI shows up everywhere, across marketing, finance, operations, procurement, network teams, and customer support. Not only does everyone want in, but the reality is that AI is already being used – both officially and unofficially – throughout organizations.

As Rick Hamilton explained: “The AI phenomena is different than what we’ve seen in the past. If you think about migration to IP, or the movement to cloud or to 5G, the industry spent the first two or three years trying to work out what it’s all about. But with AI, everybody’s in it from day one.”

But this accessibility breeds chaos unless it can be accompanied by cultural readiness. Many telcos still struggle with rigid silos, legacy mindsets and cultures of cautious experimentation. AI demands something different. It requires a “test-and-learn” mindset, where pilots and proofs of concept are encouraged but aligned to business outcomes. And, crucially, a shift from ‘command-and-control’ to co-creation, where teams help shape AI use cases rather than have them handed down.

2. Top-down vision + bottom-up execution = AI momentum

One of the strongest themes to emerge from speaking with telecom leaders is the need for a push-pull governance model. Pure top-down directives are rarely effective in isolation. Or, as VMO2’s Tadayon succinctly put it, “The system fights back.” Traditional telcos, especially those shaped by years of infrastructure-centric operations, are deeply resistant to sudden change.

For senior representatives speaking at FutureNet World from CSPs including VMO2, Vodafone and Swisscom, this means establishing mechanisms for bringing disparate, cross-functional teams together and creating a common set of goals and programs.

As Glenn Dale, Head of UK Digital and Global Head of Cloud Engineering at Vodafone, put it: “AI is a technology, but it’s not just the technology department that’s going to make this work… [CSPs need to] get all teams using AI in safe, innocuous ways and it soon gets everyone’s creative juices going!”

3. Skills: The most underestimated barrier

This is why skills – and upskilling all employees – can’t be forgotten. In the rush to deploy AI tools, many CSPs overlook the hard work of reskilling and upskilling their people. It’s not just about training data scientists. Every function, from operations to marketing, will require AI literacy.

For VMO2, skills and capability, both now and in the future, are the number one challenge, according to Tadayon.

CSPs need to focus on three skill layers:

• Technical fluency: Data engineers, cloud architects, model trainers. These are foundational roles that need to scale. CSPs are rich in infrastructure skill sets; upskilling these into a more digital software engineering mindset is a vital step towards becoming AI-native.
• AI translators: People who bridge business needs and AI capabilities. These “bilinguals” are often the difference between a failed pilot and a scalable product.
• Ethical and empathetic leadership: AI adoption also raises questions around trust, bias and job displacement. Leaders must guide these discussions with transparency and care.

4. Getting the AI strategy right

One of the most expensive lessons from the cloud era was the failure of “lift-and-shift” strategies – migrating systems without rethinking them.

That mistake risks being repeated with AI.

Rather than layering AI on top of existing workflows, CSPs need to rethink how work gets done. As Tadayon noted, real change requires understanding the interdependencies within processes and then designing AI solutions that integrate across silos.

For example, solving a “cost of faults” query might involve accessing data from network telemetry, ERP systems, and field operations. “All of a sudden, to solve that problem, you need to think: where else is my data, and how do I make it accessible?” she explained.

It’s not enough to try and use AI to just improve workflows incrementally. AI should be used as a catalyst to redesign them and embrace new cross-functional ways of working.

Against backdrop of technology chaos, fragmented teams and AI fluency, one of the most persistent blockers to AI adoption in telecoms is misaligned incentives. Different departments have different KPIs, and AI initiatives often challenge these by design. For example, a network optimization tool might improve overall efficiency, but threaten a team’s control over their own processes.

The solution lies in shared KPIs that reflect collective ownership. When marketing, engineering and operations are all measured by customer churn or fault resolution time, alignment naturally improves. A cross-functional “we” mentality starts to take root.

Conclusion

For telcos, the road to AI-native operations starts not with tools, but with transformation. This means leadership clarity, a reskilled and upskilled workforce and cultural openness.