Photo by Qeis Ismail on Pexels
Telcos Eye a New Revenue Frontier as AI Demand Explodes
The telecommunications industry has spent years searching for its next great revenue opportunity beyond connectivity. Voice revenues have eroded, SMS has been cannibalized by over-the-top messaging apps, and the promised riches of 5G have been slower to materialize than many operators hoped. Now, a new partnership between cloud-native network software specialist Mavenir and open-source enterprise technology giant Red Hat is aiming to hand telcos the tools they need to stake a serious claim in the artificial intelligence economy.
The collaboration centers on enabling telecom operators to expose, manage, and — critically — monetize AI services delivered across their networks. At the heart of the solution is a token-based charging mechanism that allows operators to bill for AI usage in ways that closely resemble the metered, usage-based models already deeply embedded in telecom billing infrastructure. For an industry that knows how to charge for minutes, megabytes, and messages, the transition to charging for AI tokens is a conceptually elegant leap.
What Is Token-Based Charging and Why Does It Matter?
In the world of large language models and generative AI, a “token” is the fundamental unit of computation — roughly corresponding to a word or word fragment processed by an AI model. Major AI providers like OpenAI already charge developers per thousand tokens consumed via their APIs. Mavenir’s insight is that telcos, sitting atop vast network infrastructure and existing billing relationships with millions of enterprise and consumer customers, are uniquely positioned to act as intermediaries and value-added resellers of AI capabilities using this same unit economics model.
By integrating token-based charging into the operator’s existing online charging system (OCS) or converged charging infrastructure, telcos can track AI service consumption in real time, apply policy controls, enforce spending caps, and generate itemized bills — all capabilities they have refined over decades of managing complex mobile and broadband service plans. This lowers the technical barrier for operators who might otherwise struggle to build AI monetization stacks from scratch.
Red Hat OpenShift as the AI Infrastructure Foundation
On the infrastructure side, Red Hat brings its OpenShift container platform to the table, providing the cloud-native runtime environment on which AI workloads and network functions can be co-hosted and orchestrated. OpenShift’s role is significant: it offers telcos a standardized, enterprise-grade Kubernetes environment that supports both the existing virtualized network functions (VNFs) and cloud-native network functions (CNFs) operators already run, while simultaneously providing the computational substrate for AI inferencing workloads.
This convergence of network and AI infrastructure on a common platform is more than an architectural convenience. It enables operators to deploy AI models at the edge of their networks — closer to end users and enterprise customers — reducing latency for AI-powered applications and allowing telcos to differentiate their AI offerings based on network performance guarantees. The ability to offer low-latency AI inferencing backed by service level agreements (SLAs) is something hyperscalers operating from centralized data centers simply cannot match in every geographic scenario.
The Broader Telco AI Monetization Landscape
Mavenir and Red Hat are not alone in recognizing the opportunity. Across the industry, operators including Deutsche Telekom, SK Telecom, and SoftBank have been making aggressive moves to embed themselves in the AI supply chain — whether through partnerships with AI model developers, the creation of sovereign AI platforms, or the buildout of GPU-equipped edge data centers. GSMA Intelligence has estimated that AI-related services could contribute tens of billions of dollars in incremental revenue to the global telecom sector by the end of the decade.
What makes the Mavenir-Red Hat approach noteworthy is its emphasis on leveraging existing telecom-grade charging and policy infrastructure rather than requiring operators to deploy entirely new monetization stacks. This pragmatic approach reduces time to market and integration risk — two factors that have historically slowed telco technology adoption. For operators already running Mavenir’s cloud-native core network software or Red Hat’s OpenShift-based infrastructure, the on-ramp to AI monetization becomes substantially shorter.
Enterprise Customers: The Primary Target Market
While consumer AI applications will certainly play a role, industry analysts broadly agree that enterprise customers represent the most immediate and lucrative monetization opportunity for telecom operators in the AI space. Enterprises are actively seeking managed AI services that come bundled with connectivity, security, and the kind of guaranteed performance levels that consumer-grade cloud AI services cannot provide. Telcos, with their existing enterprise sales channels and network SLA capabilities, are well positioned to serve this market — provided they have the billing and service management infrastructure to support it.
The token-based charging model maps neatly onto enterprise procurement models as well. IT departments accustomed to buying cloud services on a consumption basis will find token-denominated AI billing familiar and auditable, easing the procurement process for telco-delivered AI services.
Industry Outlook: Infrastructure Providers or AI Enablers?
The fundamental question facing telecom operators in the AI era is whether they will be content to serve as dumb pipes carrying AI traffic generated by others, or whether they will successfully insert themselves into the AI value chain as active service providers and monetization layers. Partnerships like the one between Mavenir and Red Hat represent a concrete, technically grounded attempt to enable the latter outcome.
As AI inference workloads continue to migrate toward the network edge and as enterprise demand for managed AI services grows, operators that have built the billing, orchestration, and infrastructure capabilities to support AI monetization will find themselves with a meaningful competitive advantage. The window to establish that position is open now — but it will not remain open indefinitely as hyperscalers and new entrants sharpen their own edge AI strategies. For telcos willing to act decisively, token-based AI charging may prove to be one of the most valuable new revenue mechanisms since the dawn of mobile data.
