With their vast network of real estate, extensive fiber connections, and established power infrastructure, telecommunications companies (telcos) are uniquely positioned to repurpose their underutilized central offices (COs) into hubs for artificial intelligence (AI). This strategic move can create new revenue streams, optimize network performance, and position telcos as key players in the growing AI economy.
Most of us can remember those boring looking red brick buildings with no windows that had a sign affixed “Bell Telephone Co”. Traditionally, those central offices served as critical nodes for routing voice and data traffic. However, with the shift to cloud-based services and more efficient network technologies, many of these facilities are now underutilized. Over the past ten years many of these properties were sold to save expenses on power, and real estate taxes. Many were located in choice urban or suburban areas, and the telcos took advantage of rising real estate prices when they wanted to shed expenses.
Instead of selling off these valuable assets, some operators have decided to transform them into "AI factories" or edge data centers. These repurposed COs could house the powerful compute infrastructure—specifically, high-performance GPUs—needed for AI model training and inferencing.
Why are these assets so valuable?
- Proximity to Customers: Repurposing COs for AI creates a distributed network of edge data centers that are physically closer to end-users and businesses. This proximity is crucial for running latency-sensitive AI applications, such as real-time analytics for autonomous vehicles, industrial robotics, and augmented reality.
- Existing Infrastructure: Telcos already have the necessary infrastructure in place. This includes fiber optic networks for high-speed connectivity, existing power grids, and physical security measures. Upgrading these sites is significantly faster and more cost-effective than building new data centers from the ground up.
- New Revenue Streams: By offering AI infrastructure as a service (e.g. Graphical processing Units GPU-as-a-Service), telcos can attract new customers, including enterprises, startups, and developers who need access to powerful computing resources without the massive upfront capital investment. This can also include providing sovereign AI services, which allow governments and organizations to develop and run AI models that comply with local data residency and security regulations.
Strategic Partnerships and Practical Examples
Telcos aren't doing this alone. Many are forming strategic partnerships with technology companies and cloud providers to build and scale these new services.
- SoftBank and NVIDIA: Japan’s SoftBank is working with NVIDIA to build AI-RAN (Radio Access Network), a new type of wireless network that integrates AI processing directly into the RAN. This partnership aims to repurpose SoftBank’s existing infrastructure for AI workloads, enabling low-latency, localized AI services like remote support for autonomous vehicles and factory monitoring.
- Verizon and AWS: Verizon has partnered with Amazon Web Services (AWS) to offer edge computing services. By leveraging Verizon's 5G network and central offices, this collaboration brings AWS cloud services closer to the end-user, significantly reducing latency for applications that require real-time data processing.
- Global Telco AI Alliance: Leading telcos, including SK Telecom, Deutsche Telekom, e&, and Singtel, have formed an alliance to develop and commercialize a telco-specific large language model (LLM). This collaborative effort leverages the collective data and infrastructure of its members to create AI solutions tailored for the telecommunications industry, from enhancing customer service with AI-powered chatbots to optimizing network operations.
The Broader AI Opportunity for telcos
The opportunity for telcos extends beyond simply renting out space and power. By leveraging AI, they can also improve their own internal operations and create new services.
- Network Optimization: AI and machine learning can analyze network data in real-time to predict and prevent outages, optimize traffic flow, and reduce energy consumption. For example, AT&T uses AI to design, build, and maintain its networks, making decisions about cell site locations and capacity planning with greater efficiency.
- Enhanced Customer Experience: AI-powered chatbots and virtual assistants can provide 24/7 customer support, while AI can analyze customer data to offer personalized product recommendations and resolve issues more quickly.
- Innovative Services: Telcos can use their AI infrastructure to offer new, value-added services to businesses, such as a GPU-as-a-Service platform for AI developers or a platform for building private 5G networks with integrated AI for industrial applications.
By strategically repurposing their central offices, telcos are not just adapting to a changing market; they're actively shaping the future of AI infrastructure. This transformation turns a potential liability—underutilized real estate- into a significant competitive advantage, paving the way for a new era of growth and innovation.