Enterprises exploring generative AI are finding that model training, inference, and data movement place heavier demands on networks, creating new challenges around latency, cost, and compliance for technology leaders. Verizon Business is positioning its network services to address these changes.
Under a new agreement with AWS, Verizon Business will add long‑haul fibre routes linking key AWS data centres. These dedicated links are intended to provide organisations with more reliable paths for AI workloads that rely on live data streams, shared training jobs, and cloud‑based toolsets.
This initiative is part of a broader response to rising network demand. Training advanced AI models already consumes substantial compute resources, but the shift toward real‑time decision‑making and live inference is expected to increase pressure on networks. Analysts anticipate many AI tasks moving toward live inference by the end of the decade, which will require faster connectivity, expanded edge capacity, and tighter security controls. These trends will affect how organisations design systems, budget for infrastructure, and handle data governance.
To support this transition, Verizon Business has launched Verizon AI Connect, a suite of network services tailored for AI workloads. Rather than building entirely new infrastructure from scratch, the company is reworking and optimizing parts of its existing network—expanding 5G coverage, reinforcing fibre routes, growing edge compute capacity, and improving cooling and power for large‑scale processing. Major cloud providers and hyperscalers are already leveraging this additional capacity to meet rising demand.
The benefits extend beyond simply adding faster cables. For example, a bank trying to detect fraud in real time can reduce false positives and speed decision cycles by moving large datasets between systems more quickly and reliably. Improved network performance can also shorten model training cycles, enable rapid testing of ideas, and prevent AI workloads from interfering with other critical business applications that share the same network resources.
Partnerships across the industry reflect how the space is evolving. NVIDIA is collaborating with Verizon on GPU‑based platforms that can operate within private 5G networks, enabling customers to keep sensitive workloads on site. Vultr is teaming with Verizon to offer GPU cloud access for on‑demand training and inference. Google Cloud is working with Verizon on AI systems designed to detect network issues before they impact customers. Verizon’s collaboration with Meta aims to extend optimisations deeper into the network to support that company’s AI initiatives.
Past work between Verizon and cloud partners shows that performance gains depend on coordinated effort rather than any single tool. Real improvements come when network engineers, data teams, and application developers collaborate to plan traffic paths, set priorities, and monitor usage. Many early AI deployments stall when teams overlook fundamentals such as data inventory, skills training, or alignment with existing security and compliance controls.
Nevertheless, obstacles remain. Organisations need better visibility across multiple network paths to understand where traffic flows and how workloads are routed. They must balance the cost of transferring large datasets between cloud regions against performance and latency requirements, and ensure routing choices comply with regulatory constraints. AI workloads can surge unpredictably, complicating capacity planning. Security teams must validate that monitoring, encryption, and access controls remain effective under higher volumes and distributed architectures.
Interest in moving AI closer to users and devices reflects a broader shift to reduce latency and increase stability. That movement is driving tighter collaboration among network operations, cloud architects, and security teams, particularly as compliance regimes evolve. In industries where even small delays matter—finance, healthcare, logistics—the underlying network performance will increasingly influence efficiency, user experience, and competitive positioning.
As AI adoption grows, the structure and management of network infrastructure will play a central role in determining long‑term cost, operational resilience, and the quality of AI‑driven services. Organisations that invest in optimized connectivity, edge capacity, and cross‑discipline coordination will be better positioned to scale AI while managing latency, security, and compliance risks.
(Photo by Leon Bredella)
Want to discover how IoT is transforming telecoms and connectivity? Attend the IoT Tech Expo events in Amsterdam, California, and London to explore innovations in 5G, edge computing, and IoT that are shaping the future of networks and services. The event is part of TechEx and is co‑located with other technology conferences.
Telecoms News is produced by TechForge Media. Explore upcoming enterprise technology events and webinars through TechForge Media’s event listings.