The evolving role of service providers—from connectivity pipes to AI-fabric providers

The second analyst views in this three-part series explores how the emergence of AI presents telecom service providers with a strategic opportunity to evolve beyond the traditional "dumb pipe" connectivity model.

If you miss our first analyst views “From choked pipes to AI platforms: The CUPS–metro monetization path”, click here: https://omdia.tech.informa.com/blogs/2026/july/from-choked-pipes-to-ai-platforms-the-cups-metro-monetization-path

The rise of AI presents telecom service providers with a strategic opportunity to evolve beyond traditional connectivity—from "dumb pipe" providers to AI-fabric enablers. By embedding AI across their network connectivity, computing, and service layers, they can transform their  IP transport network infrastructure from a cost center into a revenue-generating asset. This shift to AI-fabric providers is now essential and strategically vital for survival in the AI-driven economy.  

AI presents a new growth opportunity for telcos, much like the historic bundling of telco triple-play or quad-play revenue services, clearly signaling a business opportunity for all telco providers beyond merely providing connectivity. However, it’s time for telcos to learn from their past experiences of losing market share to cloud computing providers, content and media providers, and fintech providers, despite owning essential connectivity infrastructure and high-bandwidth service offerings. Telcos generally have a slightly easier path to becoming AI-fabric providers than hyperscalers or cloud companies, mainly because they already possess comprehensive end-to-end network infrastructure—from last-mile fiber to underlay IP, optical networks, and data centers. AI naturally positions telcos as strong competitors in the AI economy. 

However, capturing this opportunity requires more than existing assets. To succeed, telcos must:

• Understand AI workloads and incorporate AI networking capabilities by modernizing their IP transport networks. This role transformation from connectivity to AI-fabric provider involves shifting their operating model toward network architectural simplification.

• Modernize IP transport networks by implementing AI-optimized routing, advanced 400GE/800GE routers, and edge or distributed compute capabilities.

• Adopt AI-powered operations, such as AIOps for autonomous, self-healing networks. 

• Support multi-layer awareness to deliver differentiated, premium connectivity services with over 99.99% uptime and under 10ms latency for AI traffic.

The competitive landscape: A window, not a guarantee

In a nutshell, the AI infrastructure economy is emerging, but the ultimate winners have not yet been decided. Hyperscalers and cloud providers are continuously expanding and upgrading their data centers and cloud services. However, they do not control the end-to-end IP transport network that is essential for an AI-fabric provider. 

Partnering with telecom companies is a must for hyperscalers to achieve ultra-low latency. To lead the market, telecom companies need to shift from their traditional, slow, and cautious approach to adopting new technologies. 

They should capitalize on their existing connectivity expertise and extensive experience in managing complex networks to position themselves as genuine AI-fabric providers.