Looking Forward to MWC Barcelona 2026: Intelligent Infrastructure in the IQ Era

This year’s Mobile World Congress (MWC) Barcelona on March 2–5, 2026, at Fira Gran Via feels especially anchored around one question: what does “AI everywhere” actually demand from the infrastructure underneath it?

The MWC26 program is organized around six thought-leadership themes: Intelligent Infrastructure, ConnectAI, AI 4 Enterprise, AI Nexus, Tech4All, and Game Changers, which makes it easy to get pulled in many directions at once. For our clients, the most useful conversations tend to be the ones that connect AI ambition to real operational constraints: latency & performance, cost control, resilience, and the practical realities of managing hybrid network infrastructure. Through that lens, here are four sessions we’re watching closely in the official agenda.

Manufacturing and Production Summit: Building AI-Ready Networks for Resources

One session we’re circling early in the week is Manufacturing and Production Summit: Building AI-Ready Networks for Resources. The pitch is refreshingly grounded: the resources sector is aiming for around 90% automation over the next 2–3 years, but a lot of teams are still wrestling with the same reality: technical debt, disconnected systems, and not enough room to innovate for fear of downtime. So the focus here isn’t “let’s add AI” or “here’s how AI is going to be incredible” as much as “let’s make the network and the operating model capable of supporting AI at scale.”

The agenda description leans into the idea that the key is in next-gen network capabilities: connecting thousands of assets, pulling front- and back-office systems into a more unified flow, and building advanced control centers that can run predictive workflows and drive “zero-waste” performance. That all sounds right, and it’s also exactly the part that tends to get messy in the real world, especially with environments where you don’t get many chances to break things “just to modernize.”

What we’re really hoping to hear is how the speakers translate “AI-ready” into decisions you can make and plans you can implement. The session calls for a seamless IT/OT environment anchored to a unified core, plus modernized infrastructure and robust security frameworks – which we absolutely agree with, but has historically been the most elusive to implement.

Here’s what we’ll be listening for:

• Where does that “unified core” actually live? Cloud, edge, on-prem, or hybrid and why? What gets processed locally, what goes central, and how do they track that?
• How do you deal with technical debt without freezing progress? What gets refactored vs. wrapped vs. ripped out, and how do teams decide quickly?
• How do control centers actually run day to day? What’s genuinely automated, what stays human-in-the-loop, and what resilience looks like when something goes sideways.

If the conversation gets into the gritty details, like who owns what, what the rollout looks like, what surprised them, what they’d do differently, then this could be one of the more useful sessions for anyone trying to move industrial AI from pilots to something durable in production.

From network operations to network intelligence

One of the “state of the union” checkpoints on the infrastructure shift is Connected Intelligence: Building the Intelligent Infrastructure for the AI Era. Sponsored by Snowflake, the framing of this round-table discussion is pretty direct: competitive advantage is moving from “having connectivity” to “turning data into automation and value,” with operators, vendors, and connected ecosystems “building the data foundations for AI-native networks.

With the first key discussion point being about how telecoms are transforming network architecture capable of supporting AI at scale, what we expect to hear is more about the foundations: how the platforms are governed and scaled, real-time analytics, and automation that works with and for operators in their production networks.

We’ll also be listening for:

• Where do teams really start when they say “AI-native networks”? Data model? Observability? Identity? Event pipelines?
• How are leaders thinking about understanding and controlling an environment of “shared intelligence”, and how are they planning to use the information once they get it?
• Which automation use cases are delivering measurable reliability gains today and which are still aspirational?

Data centers as the AI constraint, not just the AI enabler

For those of us who care about the cost of delivering payloads (as we certainly do) Cloud Fusion: Mastering Data Centers for the AI Era stands out as an explicit “ops meets AI” session. The description focuses on AI-augmented data centers improving operational efficiency through intelligent resource management, predictive analytics, and on-demand scalability while also calling out environmental impact as a first-order design constraint.

The value here is practical: if AI is becoming an always-on workload, then data center operations become a competitive function, not background infrastructure – and the cost of managing them becomes of paramount importance. It’s also one we’ve seen before, being reminiscent of the virtualization/cloud utilization battle 20 years ago. As a new technology scales up, the race to make it more efficient and get more value out of the same input becomes a differentiator. If companies can view AI efforts more strategically, there is value in keeping costs in check, but also getting faster responses with better outputs.

The agenda’s emphasis on automation and predictive analytics suggests we’ll hear about capacity planning, workload placement, and the reality of supporting everything from 5G workloads to heavier AI compute without fragilizing the environment.

What we’ll be listening for:

• How leaders manage utilization when workloads are spiky, multi-tenant, and model-driven (and when underutilized accelerators become a silent tax).
• How organizations are balancing performance targets with energy constraints and sustainability goals.
• What “AI-augmented operations” looks like day-to-day: which signals matter, what gets automated, and what still requires human judgment.

Looking forward

Taken together, these four sessions map a familiar priority arc: data foundations → operational automation → economics and monetization → physical and energy constraints → the next-gen architecture stack. MWC26’s theme structure reflects the industry’s desire to connect those dots.

As we head into Barcelona, what we’re hoping for most is that the conversations move past polished narratives and into the practical: what’s working, what’s brittle, what teams are standardizing, and what they’re walking away from. If the speakers get specific about trade-offs, like cost vs. control, central vs. edge, and openness vs. differentiation, we’ll come away with signals our clients can actually use.