MWC26 Day 1 Recap
No surprise – AI is everywhere at MWC. As ever, it’s hard to separate real value from buzzword bingo.
Today the Cellusys team focused on network evolution rather than AI (we’ll cover AI in more detail in the coming days). In reality, almost every discussion around 5G, 6G, and network evolution leads to AI. Fair enough, everyone wants a share of AI’s potential. The issue is that many discussions feel driven more by hope than by a clear, practical plan. So, what else stood out today?
Sovereignty
‘Sovereignty’ was a word we heard repeatedly in the sessions we attended. Not the most exciting theme perhaps; but refreshingly grounded in reality compared to the AI hype surrounding everything else.
The first session I attended focused on Europe, with strong contributions from Orange and Airbus. The broader point was clear: states and blocs are increasingly retreating behind their own walls, each with their own security and privacy regulations, and those regulations are becoming more significant. The panel highlighted the need for technology, infrastructure, and data sovereignty. Over-reliance on largely American internet giants is a concern, particularly as regulation diverges across regions. We can expect countries and blocs to invest more in their own infrastructure and technology. It’s a major undertaking, but increasingly seen as necessary to ensure security and privacy.
Edge Computing and Efficiency
Edge computing was heavily touted as part of 5G, but never fully materialised. It was largely framed around low latency and use cases like autonomous driving. A different perspective emerged: the issue may not just be latency, but reliance on the cloud.
- AI will generate far more data; it may simply be inefficient to send all of it to the cloud.
- From a data sovereignty perspective, sending sensitive data to external cloud providers is not always desirable.
More broadly, efficiency — at the edge or in the cloud — was positioned as critical in an AI-driven world. AI is resource-intensive, and systems must be designed accordingly. The era of cheap, abundant resources is fading. Bigger is no longer better. Efficiency is. This was particularly clear in discussions around non-terrestrial networks. In space, you cannot just add more resources to solve the problem.
5G
As ever, discussion around 5G centred on monetisation and real-world use cases. The tone in the sessions we attended was more positive than expected, though it still feels closer to hope than widespread reality.
NUHS and Elmo shared interesting examples in medical and remote driving scenarios that rely on network slicing. In Elmo’s case, the requirement was not raw speed, but assured service. 5G may not be the only way to deliver that, but slicing does provide a workable approach.
Slicing has always been the most compelling and potentially disruptive aspect of 5G, yet progress has been slow. Some credible use cases are emerging, as we saw today. Still, whether 5G will deliver value at scale remains an open question. Despite the strong presentations, I remain cautious on scale.
6G
Discussion centred on what 6G should become. Two themes stood out:
- 6G networks should be ‘AI-native’
- Seamless operation between terrestrial and non-terrestrial networks
The second point is more tangible. Quentin van de Geer of Airbus made strong arguments around resilience. Different network types working together can improve robustness. One example was optical communications that are immune to RF jamming. As with sovereignty, resilience was a recurring theme, and it makes sense to embed this into 6G thinking.
The ‘AI-native’ concept was less concrete. Operators see it as a way to create new value and move beyond the “dumb pipe” model, offering AI-powered networks rather than simply carrying traffic for OTT AI providers. Detail was understandably limited (6G is still distant), but it does feel like an area operators are attaching future value to. The direction is logical. Execution is another matter. We are not far removed from the promised remote surgeries and autonomous driving of 5G. Avoiding another cycle of hype without substance will be critical.
Perhaps the most grounded point came from Bruno Zerbib of Orange. He noted that defining a 6G specification today and expecting it to remain relevant into the mid-2030s is unrealistic. The real requirement is not another ‘G’, but a more flexible and agile operating model.
Operators’ strongest asset is trust. If operators can build adaptable solutions around customer needs, they can reinforce that position. But this will not come from a static 6G blueprint, it requires deeper changes in how they operate.
Day 2 Watchlist
More tomorrow. Our main focus will be security, from fundamentals to emerging areas such as quantum. And, as ever, we’ll report on whatever else stands out along the way.
Daniel McTague, CTO, Cellusys
Categorised in: Blog