Why Swisscom’s CTIO is focused on data model development

• AI is helping telcos to break new operational ground 
• Mark Düsener tells the FutureNet World audience that it took just two people for Swisscom to develop a radio access network (RAN) digital twin using AI 
• Such a task was unimaginable just two to three years ago
• But using AI at scale and to its full potential requires the development of a unified data platform, noted Swisscom’s CTIO

The ability of just two people using AI to develop a useful radio access network (RAN) digital twin, a task that would have been “unimaginable” just a few years ago, shows just how much impact AI can have on telco operations, Swisscom’s CTIO Mark Düsener told the audience at FutureNet World 2026 in London last week. But he cautioned there’s still a lot of work to be done on data management in order for the full power of AI to be unleashed. 

Düsener noted during the panel session ‘The Future Network: What Does the Network of the Future Look Like in an AI World’ that Swisscom has identified building a unified data platform as a top priority to facilitate the Swiss operator’s switch to AI-powered networking. He explained that although AI is already having a notable positive impact on operations, the operator has realised that AI cannot scale without a strong data platform that allows engineers to query the current network state in seconds directly from a real-time data stream.

In the meantime, AI is enabling significant advances. Düsener told the crowd at the London event how a student doing a master’s thesis at the telco had, with minimal support, built a digital twin of the operator’s RAN that can predict energy savings better than any existing tool.

“This student, [with just] one of our colleagues helping him, built a digital twin of our RAN. This meant we could predict energy savings much better than using any tool we had so far. That shows the potential of AI in the network – solving a problem at faster speeds with lower resources, to an extent we couldn’t [have] imagined two or three years ago,” noted the CTIO. 

“The most interesting thing is that it took just two people [to achieve] something that a few years ago was the Holy Grail,” he explained.

According to Düsener, the story also highlights the importance of data, something that network operators have in abundance and which can be put to great use. Telcos own “tremendous amounts of data”, the Swisscom CTIO noted, producing billions of data points per day. “Turning that into knowledge is the goal,” he added.

To make the most of its data assets, Swisscom is building a data model that ingests source data and adds context that can be used across the organisation by anyone using natural language prompts. “The basis of all AI is data. At the start of the year, my team asked me what was priority number one for all the transformations we are doing, and that is it,” he added.

“If we want to know how a product is performing, or how a customer is behaving, we can just use the data, and I believe this is one of the most transformational things we can do,” stated Düsener.

Also on the panel were Andrea Donà, chief network officer at UK operator VodafoneThree, Gabriela Styf Sjöman, managing director of research and networks strategy at BT Group, and Joe Hogan, networks CTO At Amdocs.

Donà agreed that data governance is vital, but the debate remains as to whether organisations should appoint an individual to lead this, such as a chief data officer, or whether it should be a domain-centric role, with the VodafoneThree exec preferring the latter.

“It needs to be domain centric because you cannot lose the context of that particular domain,” he explained. “But to do that to a high capability, you need to start simplifying your OSS. Before you talk about the data you need to talk about OSS simplification.”

BT’s Styf Sjöman discussed the operator’s focus on intelligent operations, sustainability and cyberdefence, noting that BT’s approach is based on the “network immune system” concept, inspired by biological systems. The initiative uses AI for autonomous containment, red teaming and recovery from cyber threats, such as adversarial AI phishing attacks.

“We need to use AI to fight AI. Today, 80% of phishing attacks are AI generated, and these attacks are fast, they are cheap, and they are on all the time,” she said. “We are looking at how we use AI for defence and response, as well as protection. This goes hand in hand with intelligent automation because resilience is about using AI for recovery.”

She added that BT takes a lot of lessons from biology and how the body responds to attacks, using the network as a form of an immune system, with AI leveraged as part of autonomous containment.

- James Pearce, Editor, TelecomTV