Solving a Commercial Problem, Not a Technical One
On June 24, Beyond Now announced an expansion of its Wave AI Framework, integrating it with Microsoft Foundry so communications service providers can build, deploy, package, bill for, and monetize agentic AI services. What makes this announcement worth attention is not the technology pairing, which is straightforward, but the problem it names. The framing is refreshingly honest: telcos have plenty of AI pilots, but not enough repeatable ways to sell them. That is the unglamorous truth across the telecom sector and well beyond it. The hard part of enterprise AI in 2026 is no longer building a working model, it is turning a working model into a product someone will pay for, repeatedly, at scale.
Beyond Now CEO Angus Ward put the thesis plainly, saying AI agents will only create meaningful value for CSPs if they can be connected to real business processes, customer journeys and monetization models. We think that sentence applies far beyond telecom. The graveyard of enterprise AI is full of impressive demos that never found a billing model, an owner, or a place in an actual customer journey. By foregrounding monetization rather than capability, Beyond Now is implicitly criticizing the entire pilot-obsessed phase of enterprise AI and arguing that the next phase is about commercial machinery. That is the correct read of where the market is heading.
Microsoft Supplies the Engine, Beyond Now Supplies the Storefront
The division of labor is clean. Microsoft supplies the agent-development substrate through Foundry, the place where the agents get built and run. Beyond Now supplies the business machinery around it: the Agentic Hub, the marketplace, the billing, and the ecosystem orchestration. Rick Lievano, Microsoft's CTO for telco, media and gaming, said the approach enables CSPs to accelerate the adoption and monetization of AI while delivering scalable, secure solutions built on Microsoft Azure. In practice this means a service provider can activate, orchestrate, package, sell, and monetize both Microsoft and third-party AI agents from a single platform, rather than stitching together a development tool, a billing system, and a marketplace by hand.
This layering is strategically significant for how value gets captured in the agent economy. Microsoft owns the substrate and the cloud, which is a strong position, but the company is conceding the commercialization layer to a specialist partner rather than trying to own everything. For Beyond Now, sitting between the model platform and the end customer's billing relationship is a deliberately defensible spot. Whoever owns the marketplace, the packaging, and the metering owns the customer relationship and the recurring revenue. We have watched this pattern play out across enterprise software repeatedly, and the lesson holds: the storefront and the billing engine are often a better business than the underlying technology.
Telcos Want to Be More Than Pipes Again
There is an old anxiety underneath this deal that every telecom executive will recognize. Carriers have spent two decades watching internet companies build enormously profitable businesses on top of the connectivity they provide, while their own role compressed into being a commodity pipe. Agentic AI presents both a fresh version of that risk and a chance to escape it. Beyond Now's pitch is explicitly that CSPs can become more than connectivity wholesalers in someone else's AI boom. By giving carriers a way to package and sell agents, the framework offers them a path to capture some of the AI value flowing across their networks rather than merely carrying it.
Whether telcos can actually execute on that ambition is a separate question, and the honest answer is that the sector's track record on moving up the value chain is mixed. But the strategic logic is sound, and the tooling gap was real. A carrier that wants to sell AI services to its enterprise customers has the relationships and the trust, what it has lacked is the commercial plumbing to turn a pilot into a repeatable, billable product. Beyond Now and Microsoft are trying to supply exactly that plumbing. The opportunity is meaningful: CSPs sit on enormous enterprise customer bases and the operational data to make industry-specific agents genuinely useful, if they can package them fast enough.
The Monetization Layer Is the 2026 Story
Zoom out and this telecom-specific deal is a clean illustration of a market-wide shift. Across enterprise software, the conversation in 2026 has moved from whether agents work to how anyone gets paid when they do. Salesforce bought a usage-based billing specialist to bring consumption pricing into Agentforce. ServiceNow, SAP, and Workday are building metering layers so that agents calling their systems generate revenue. Now Beyond Now is doing the equivalent for telcos. The common thread is unmistakable: the monetization layer is where the next phase of enterprise AI competition is being fought, because the technology is increasingly commoditized while the commercial model is not.
For CIOs and strategy leaders, the takeaway is to watch the billing and packaging architecture as closely as the model capabilities. The vendors quietly building marketplaces, metering engines, and consumption pricing are positioning to own the customer relationship in the agent era, and that ownership is sticky. Beyond Now's June 24 expansion will not make headlines the way a frontier model release does, but it is arguably more representative of where the real money and the real lock-in are forming. The pilots were the easy part. Turning them into a repeatable, billable, governed product is the work that separates the winners from the long list of impressive demos that never paid for themselves.
