The Nervous System of the AI Economy
Why telecom infrastructure is becoming the strategic layer of artificial intelligence
Barcelona — Inside the vast halls of the Fira Gran Via exhibition centre, the annual gathering of the global telecom industry unfolds with familiar spectacle. The latest smartphones promise ever more intelligent features. Startups demonstrate autonomous robots, AI-powered assistants and connected devices designed for the next phase of the digital economy.
Yet behind the gadget launches and technology demonstrations at Mobile World Congress 2026, a quieter but far more strategic debate is taking place. In executive briefings and panel discussions, the conversation has shifted away from apps and devices toward something deeper: the physical infrastructure that will underpin the next generation of artificial intelligence.
For much of the past decade, the AI narrative has been dominated by software. Technology giants building ever more powerful models have captured the imagination of investors and policymakers alike. But as AI begins to move beyond chatbots into factories, logistics networks and autonomous machines, attention is gradually shifting toward the networks that enable those systems to operate in real time.
During one of the keynote sessions at the conference, Allison Kirkby, the chief executive of BT Group, summarised that shift in a remark that quickly circulated across the telecom industry.
“Telecom networks are no longer merely utilities; they are the nervous system of modern society. Our infrastructure is the foundation upon which the AI revolution rests.”
— Allison Kirkby, Chief Executive Officer BT Group
Her comment reflects a growing recognition within the technology sector: the future of artificial intelligence will depend not only on algorithms and computing power, but also on the networks capable of connecting billions of devices, machines and sensors in real time.
From the model race to the infrastructure race
The global race in artificial intelligence has largely focused on the development of increasingly powerful models. Companies such as OpenAI, Google and Microsoft have invested billions of dollars in training large-scale AI systems capable of generating text, images and software code.
But the next stage of AI development is increasingly defined by how these systems are deployed in the real world.
Training large models happens in enormous cloud data centres. Their day-to-day operation — known as inference — increasingly occurs closer to the user, often at the edge of the network.
That shift has profound implications for digital infrastructure. Real-world AI systems must process data and respond almost instantly. Autonomous machines, industrial robots and intelligent logistics systems cannot wait for information to travel across continents to distant cloud servers.
Industry experts at the conference repeatedly pointed to latency as the decisive factor.
Without ultra-low latency networks — often below ten milliseconds — AI-driven physical systems become unreliable. In extreme cases, they simply make too many errors to operate safely.
This technological reality is one reason telecom operators believe their role in the digital ecosystem may be about to change.
The end of the “dumb pipe” era
For years, telecom operators complained that they had become little more than providers of connectivity, while internet platforms captured the economic value created by the digital economy.
Companies such as Apple, Amazon and Meta Platforms built enormous ecosystems on top of telecom networks, turning those networks into what critics often called “dumb pipes”.
Kirkby argues that the AI era is changing that equation.
“The days of ‘dumb pipes’ are behind us. We are building an intelligent platform that not only transports data, but understands and prioritises it.”
— Allison Kirkby, Chief Executive Officer BT Group
Under her leadership, BT has accelerated its fibre rollout across the United Kingdom, aiming to connect 25 million homes and businesses by the end of 2026. At the same time, the company is using artificial intelligence internally to automate operations and simplify the complex legacy systems accumulated during its 180-year history.
AI, Kirkby has suggested, may even reshape the telecom workforce itself.
“Depending on what we learn from AI, there may be an opportunity for BT to become smaller and more efficient by the end of the decade.”
— Allison Kirkby, Chief Executive Officer BT Group
Such comments highlight a broader transformation within the telecom industry, where companies are increasingly positioning themselves not simply as connectivity providers, but as digital infrastructure platforms.
AI-native networks: the next technological leap
One of the most important developments discussed at the conference is the emergence of AI-native network architecture.
A coalition including NVIDIA, Ericsson, BT Group and T-Mobile is working to integrate artificial intelligence directly into the core of mobile network infrastructure.
These efforts focus particularly on the Radio Access Network (RAN) — the component of the mobile system that connects devices to the wider network.
Instead of static configurations controlled by hardware, AI-native networks continuously analyse traffic patterns, predict demand and dynamically allocate resources.
Kirkby described the shift during a discussion about the collaboration with NVIDIA.
“Connectivity is the backbone of economic growth. By building on open and trusted AI-native platforms, we are laying the foundations for an ecosystem that is intelligent, sustainable and secure.”
— Allison Kirkby, Chief Executive Officer BT Group
The implications are significant. Networks capable of optimising themselves in real time could support entirely new categories of AI-driven applications while dramatically improving efficiency.
These technologies are also expected to form the foundation for 6G, the next generation of wireless connectivity.
Physical AI and the next generation of networks
While generative AI continues to dominate headlines, many executives at Mobile World Congress are already looking toward the next phase: the integration of artificial intelligence into the physical world.
This concept — often referred to as physical AI — involves billions of connected devices making autonomous decisions in real time.
Autonomous vehicles navigating cities, robotic manufacturing systems coordinating production lines and sensor networks monitoring infrastructure all depend on constant communication between machines and AI systems.
As one panel discussion on AI-RAN technology summarised the challenge:
“6G is not about faster internet for smartphones. It is about creating a fabric for physical AI — connecting billions of machines that must make decisions in real time.”
— Panel discussion, AI-RAN Alliance session, Mobile World Congress 2026
In such an environment, the network itself becomes an integral part of the computing system.
Infrastructure as geopolitics
The strategic importance of digital infrastructure is also reshaping geopolitical debates.
Artificial intelligence is increasingly viewed not only as a technological breakthrough, but as a determinant of economic power and national security.
The United States currently dominates the AI model ecosystem through its cloud providers and software platforms. China has invested heavily in vertically integrated digital infrastructure.
Europe occupies a more complex position. Although it lacks hyperscale AI companies comparable to Silicon Valley’s giants, it remains home to several of the world’s largest telecom operators and network equipment manufacturers.
For Kirkby, this infrastructure may represent a strategic advantage.
“Digitalisation and AI have become matters of national security and economic resilience. We must manage our own infrastructure if we want to preserve strategic autonomy.”
— Allison Kirkby, Chief Executive Officer BT Group
Her comments echo a growing policy debate in Europe around digital sovereignty — the idea that nations and regions should maintain control over critical digital infrastructure.
Mapping the AI infrastructure stack
The emerging power structure of the AI economy can increasingly be understood as three interconnected layers.
| Power Layer | Dominant Players | Strategic Role |
|---|---|---|
| AI Models | OpenAI, Google, Meta | Intelligence and algorithms |
| Platforms | Microsoft, Amazon | Cloud ecosystems and access |
| Infrastructure | Telecom operators, NVIDIA, data centres | Connectivity and computing fabric |
While the first two layers have dominated public attention, the third layer — infrastructure — may prove decisive as AI expands into real-world systems.
Who controls the nervous system?
Mobile World Congress has long been a showcase for the most visible edges of technological innovation: faster networks, smarter devices and futuristic applications.
But the deeper story emerging from this year’s conference suggests a structural shift in the digital economy.
Artificial intelligence is moving from software into infrastructure.
If that transformation continues, the networks connecting machines, sensors and computing systems will become a strategic asset in their own right.
In that future, telecom operators may no longer be the quiet utilities of the internet age.
Instead, they could become the custodians of something far more powerful: the nervous system of the AI economy.
Photo / Illustration credit
Illustration: Altair Media / AI-generated (pencil sketch style)
Caption
BT Group CEO Allison Kirkby speaking at the main stage of Mobile World Congress 2026 in Barcelona, where she described telecom networks as “the nervous system of the AI economy”.
This article is part of Altair Media’s special coverage of Mobile World Congress 2026.
Follow ongoing analysis and reporting on the strategic shifts shaping global connectivity on our dedicated page:
The Future of Connectivity — MWC 2026 → https://altairmedia.eu/the-future-of-connectivity-mwc-2026/
