The 2025 Mobile World Congress didn't just spotlight flashy devices and AI demos. BT made sure the conversation included something less showy but equally foundational: networks. While the industry continues to advance with generative models, autonomous systems, and AI-infused services, BT posed a quieter question: How do we keep them running reliably at scale?
Their answer is simple but layered: the network must be treated not just as plumbing, but as part of the brain. At their booth, in panels, and through private briefings, BT consistently emphasized one point: AI implementation without strong, smart networks is like racing a Formula One car with flat tires.
The Network Isn’t Background Anymore
For decades, networks sat in the background. You sent data through them and hoped for stability. That worked when data was mostly static. Today's AI systems are real-time, data-hungry, and distributed. Generative AI must extract data from multiple sources, process it in real-time, and deliver results where needed within milliseconds. None of this works without reliable, high-throughput networks. BT highlighted this shift through several live demos and network simulations. These weren't abstract charts—they were visual cases that showed how even slight jitter or latency dropouts can cripple AI decision-making in use cases such as logistics or smart manufacturing.
BT’s CTO likened it to the nervous system in the human body—AI might be the brain, but without fast reflexes and clean signals, the whole system underperforms. For instance, if a network delay causes an autonomous vehicle fleet to update route planning late, that delay becomes more than a tech hiccup—it's a real-world disruption. AI systems rely on time-sensitive inputs. A five-second delay doesn't just slow things down—it risks wrong outcomes. BT argues that without recalibrating the entire stack to treat the network as dynamic and intelligent, these risks compound quickly.
Making the Network AI-Aware
A major takeaway from BT's presence at the Mobile World Congress was its pitch for AI-aware networks. That doesn't just mean using AI to optimize network traffic, though that's part of it. It means building infrastructure that can adapt in real-time based on the needs of AI workloads. Networks that re-route traffic before congestion happens. That isolates unstable nodes without needing human intervention. That adjusts bandwidth on the fly depending on whether a model is being trained, fine-tuned, or just queried.
BT showcased prototypes of these adaptive systems through partnerships with vendors using edge and cloud co-processing. For example, in one setup, an edge site detected a local spike in demand for a vision model tied to city surveillance feeds. Instead of sending all that raw data to the cloud, the system compressed, filtered, and partially processed it locally, then passed only the meaningful pieces upstream. This kind of flexibility cuts latency and improves reliability, especially in regions with patchy connectivity.
AI-aware networking doesn't just improve performance—it lowers cost and increases predictability. Training large models, especially in federated environments, can quickly deplete budgets if the network isn't intelligent enough to prioritize traffic effectively. BT's emphasis on "workload-driven optimization" resonated with enterprise visitors, many of whom deal with ballooning AI costs. Their demos demonstrated how you can run the same AI job faster and more cost-effectively by giving the network more responsibility.
AI in the Network, Not Just on Top of It
One of the more forward-thinking themes BT pushed was embedding AI directly into the network layers. This isn't new in theory—operators have used AI for years in traffic management and anomaly detection. What's different now is the scale and speed required. BT demonstrated how AI models, both small and large, can reside within routers and edge nodes themselves. These models detect problems, predict usage spikes, and even anticipate failure before it happens.
This distributed intelligence means the network isn’t just reacting—it’s forecasting. And that matters in sectors where milliseconds matter. In an automated factory, the ability to preempt a network slowdown could mean avoiding hours of machine downtime. In healthcare settings, an AI-aware network could dynamically prioritize patient monitoring traffic over background system logs.
BT also discussed openly how these AI-in-network systems can form the foundation for trusted AI governance. Since the network is the central point of everything flowing through it, it can serve as a checkpoint for compliance, anomaly detection, and even basic ethical controls. Think of it as an always-on traffic cop—not just monitoring what moves, but why and how it moves.
The Business Case: Scaling AI Without Scaling Chaos
Beyond the tech details, BT pushed a business narrative that resonated with enterprise leaders in the crowd. Everyone wants to scale AI. Few want to scale the headaches that come with it—ballooning cloud bills, regulatory risk, and unexpected downtime. The message from BT was that a more intelligent, integrated network can reduce all three.
They cited examples where network upgrades allowed companies to deploy AI tools to thousands of edge locations without rewriting code or changing software stacks. In another case, a logistics firm reduced failed deliveries by 12% after optimizing AI decision paths using BT’s network telemetry. This wasn't abstract theory. These were real firms with real results, and that moved the conversation beyond buzzwords.
BT also didn’t overpromise. They acknowledged limits, particularly around interoperability with legacy hardware and vendor lock-in. But their argument was clear: even if you’re not doing anything fancy with AI yet, you need to prep your network now. Waiting until the workloads arrive is like upgrading your internet mid-video call—it’s already too late.
Networks Are Now the Foundation Layer for AI
At Mobile World Congress, BT highlighted an often-overlooked part of the AI conversation: the network. AI relies on the seamless flow of data, decisions, and outcomes, all of which depend on networks built to handle the load. BT compared the relationship to a city and its roads — one cannot function without the other. Their focus was on smarter routing, edge processing, and distributed intelligence to create infrastructure ready for AI. While the event showcased impressive models and demos, BT’s point stood out: no matter how advanced the AI, it can only perform as well as the network beneath it.
