Across 40+ operator conversations, one theme has moved from experimentation to expectation: Conversational AI.
Every operator we spoke with was either actively planning for Conversational AI or quickly realizing they were behind.
When all compasses align, it is a huge step forward. And two bold announcements reinforced that sense of shared direction. Deutsche Telekom in Germany, and T-Mobile in the US, have both announced network-native, in-call AI experiences for their users.
As a reminder: T-Mobile’s Live Translation offers real-time translation in 50+ languages, and Deutsche Telekom’s Magenta AI Call Assistant includes live translation, call summaries, and in-call Q&A.
These are important signals. But they are also highly bespoke implementations—built with resources and control that most operators simply don’t have.
A route to conversation innovation
But if it’s one thing to know where you’re going, it’s something else to know how to get there, confidently and at pace. In that context, these first movers can only tell us so much.
With Magenta, you can argue that DT is acting as a hyperscaler would, internalizing a lot of the control and integration. Given the muscle it wields — market, financial, engineering and regulatory — the move makes some sense. We’ve seen this pattern before. And besides, that posture is not a luxury afforded to most operators.
More broadly our industry has to move faster and more flexibly than vertically integrated, single operator deployments.
At MWC, it was hard to walk more than a few meters without seeing another AI demo. There’s no shortage of pilots, proofs of concept, or prototypes.
Service providers have invested heavily in their networks over the past decade. Traffic is up. Complexity is up. But the revenue hasn’t followed. In some cases, when you factor in inflation, they’re moving backward.
Because there’s still a gap between ambition and execution. Most operators don’t lack ideas. They lack a repeatable way to execute them at scale.
Conversations. Experiences. Outcomes.
That’s where value is being created.
The missing layer between networks and experiences
Knowing where you’re going is only half the job. You also need a map.
That’s the gap we kept coming back to in Barcelona. The industry has a clear direction, but it’s lacking a way to get there.
Operators have battle-tested networks that run the world’s telephony. Their customers trust their actions and their intentions. On top of that, there’s a rapidly evolving ecosystem of AI-infused experiences. What’s missing is the control layer between infrastructure and experiences.
To create Magenta, DT has effectively built its own map, internalizing the control surface as a hyperscaler would.
What most operators are looking for is a horizontal orchestration layer that allows conversational AI experiences to live inside operator networks and reach global users.
Without that layer, every new use case becomes a bespoke integration. Costs increase. Time to market slows. Differentiation disappears.
With it, infrastructure becomes a platform for revenue, not just cost.
From AI pilots to P&L impact
The teams that will move fastest and furthest won’t try to do everything. They’ll start with a small number of services that consumers intuitively value, trust and use. Not just interesting use cases, but commercial ones. They’ll need to answer these questions:
How can AI make ordinary conversations extraordinary?
How does that monetise?
Will that change the economics of the business?
Enabling experiences that customers will pay for
Operators who act decisively now will define the next phase of the market. Not by doing more. By doing fewer things, with more intent and impact. By:
- Being explicit about the gap between network investment and new revenue.
- Choosing a small number of use cases that actually matter commercially.
- Being honest about the capability to execute at scale.
And by reframing the story from selling connectivity or content, to enabling the experiences that people will pay for.
The market has caught up
The biggest shift is this: service providers are no longer content to watch innovation happen around their networks. They are beginning to grasp the real opportunity to bring it into the network and turn it into accretive revenue. All compass needles point to Conversational AI. The route that operators choose will dictate who gets there in the strongest shape.
The opportunity is clear.
The technology is ready.
What happens next is an execution story.
The next 12 months won’t be defined by who talks about AI the most.
They’ll be defined by who closes the gap.
