Telecom coverage tends to treat 6G, Open RAN, and the MVNO landscape as three separate beats — infrastructure news, standards news, and business-model news, filed in different sections. That separation is misleading. Look closely at where each of these stands right now, in mid-2026, and a single connective thread runs through all three: the industry is simultaneously placing several expensive, long-horizon bets on becoming more open, more software-defined, and more AI-native — and the early results are genuinely mixed, not the clean success story any single vendor’s press release would suggest.
This is a look at where each bet actually stands, not where the marketing says it stands.
6G: the standardization reality check
Start with the timeline, because it’s more concrete than most coverage implies. 3GPP finalized the Release 21 development schedule at its June 2026 plenary in Singapore: Stage-1 freeze (service requirements and use cases) is set for March 2027, Stage-2 freeze for June 2028, with full specifications targeted for completion in early 2029. Commercial 6G systems are aimed at the 2030 window. In other words, the standard that gets discussed today as an imminent transformation is still roughly three years from being frozen, and closer to four from being commercially real.
What’s actually being decided right now is more interesting than the date itself. 3GPP confirmed key 6G air interface decisions in Singapore — waveform selection and spectrum framework parameters, built on the 5G NR foundation rather than a clean-sheet redesign. And the design philosophy is explicit: 6G is being built AI-native from the ground up, not with AI bolted on afterward. 3GPP’s SA5 working group is standardizing interfaces for AI model deployment, telemetry ingestion, and lifecycle management directly into the network architecture, aiming at closed-loop, self-governing network operation. Integrated Sensing and Communication (ISAC) — turning the radio network itself into a sensing instrument, capable of detecting and tracking objects, not just carrying data — is one of the six core use-case areas ITU-R has defined for the standard.
The market forecast attached to this — global 6G market growing from $5.2 billion in 2023 to a projected $41.4 billion by 2030 — is the kind of number that gets quoted constantly and understood rarely. It’s a bet on a standard that doesn’t fully exist yet, sized as if the outcome were already known.
Open RAN: the honest bull-and-bear case
This is where the industry narrative gets genuinely contested, and it’s worth presenting both sides rather than picking one.
The bull case is real and measurable. Rakuten Mobile’s network in Japan — the industry’s most-cited Open RAN proof point — now supports more than 5 million subscribers, reports network availability exceeding 99.5%, and has successfully integrated equipment from more than 15 different vendors across radio and core network functions. That’s not a pilot; that’s a functioning multi-vendor network at meaningful scale. AT&T’s $14 billion deal with Ericsson aims to route 70% of its US wireless traffic through open platforms by the end of 2026. Vodafone has committed to at least 2,600 Open RAN sites across rural UK by 2027, alongside expanding deployments in Germany. Government money is behind the push too — the UK has committed £250 million through its Open Networks R&D Fund specifically for this transition. Industry-wide, Open RAN represented only about 5% of the total RAN market in 2025, but is projected to reach as much as 28% by 2029 — a genuine structural shift, not a rounding error.
The bear case is just as real, and centers on one uncomfortable example. Dish Network was, for years, the American poster child for Open RAN — a fully cloud-native, multi-vendor 5G network built on AWS infrastructure with equipment from Mavenir, Altiostar, and Samsung, covering more than 70% of the US population. In 2026, Dish is dismantling that network and converting into a hybrid operator that leases AT&T’s traditional RAN under an MVNO-style arrangement, keeping only its own 5G core. The causes were largely financial and strategic, not purely technical — but the optics are unavoidable: the most ambitious Open RAN deployment in the US is now partially unwinding itself. Even committed believers are candid about the gap between hype and reality. 1&1 Mobilfunk’s CEO, discussing the company’s own greenfield Open RAN build in Germany, put it plainly at a 2026 industry panel: “The big hype… has not materialized as much as we hoped it should.”
There’s a sharper critique worth including, from telecom analyst firm Strand Consult: they argue that some major European operators are using Open RAN commitments partly as political cover — making modest Open RAN purchases to generate a favorable US-policy headline about vendor diversification, while the core of their networks remains built on Chinese equipment from Huawei and ZTE. Whether or not you accept that framing in full, it points to something structurally true: Open RAN adoption has been overwhelmingly a greenfield phenomenon — Rakuten, 1&1, Dish — while established brownfield operators, locked into 7–10 year equipment refresh cycles and wary of multi-vendor integration risk on live networks, have moved far more cautiously than the press coverage implies.
The honest synthesis: Open RAN is technically proven at scale in at least one major deployment, structurally inevitable as a growing share of new builds, and still genuinely unresolved as the default choice for the world’s largest existing operators.
MVNOs: consolidating and diversifying at the same time
The business-model layer is going through its own contradiction, one I explored in more depth after the recent European consolidation news (Liberty Global’s buyout of Vodafone’s stake in VodafoneZiggo among the most recent examples). The short version: there’s a genuine MVNO renaissance underway, but it doesn’t look like the traditional “cheap alternative” story. Fintech and retail platforms are increasingly embedding connectivity directly into their own products rather than customers seeking out a separate MVNO brand — the same logic behind Motorola building travel eSIM directly into its phone software this year. At the same time, the MVNO market itself is consolidating, squeezed from below by MNOs giving away connectivity as a loss-leader inside their own bundles, and from above by tightening regulatory requirements that favor operators with genuine infrastructure ownership (Full MVNO status) over lightweight resellers.
The connective thread: AI, not as a feature, but as the operating layer
What ties all three of these together isn’t coincidence — it’s that each is being redesigned around AI as infrastructure, not as an add-on application. 6G’s core architectural principle is AI-native design, with model deployment and lifecycle management built into the standard itself. Open RAN’s entire value proposition depends on the RAN Intelligent Controller (RIC) — a software layer that uses AI/ML to optimize radio resource allocation in real time across a multi-vendor network, something a traditional closed, single-vendor RAN never needed because one company controlled the whole stack. And the MVNOs and embedded-connectivity players surviving the current consolidation are, overwhelmingly, the ones using AI for churn prediction and personalization on top of infrastructure they either own outright or access through genuinely open interfaces.
None of these three bets works particularly well in isolation. AI-native 6G design assumes an Open RAN-style disaggregated architecture is already normal by the time 6G ships commercially around 2030. Open RAN’s promised cost and agility benefits depend on AI-driven orchestration (the RIC) actually working reliably across vendors from different companies, which is precisely the unresolved integration risk that’s kept brownfield operators cautious. And the MVNO layer’s long-term viability depends on whichever underlying network architecture — traditional, Open RAN, or eventually 6G — ends up cheapest and most flexible to build on top of.
What this means, honestly
If you’re looking for a clean verdict — “Open RAN is winning” or “6G is overhyped” — the data doesn’t support one. What it supports is a more useful, less comfortable conclusion: the telecom industry is running three interdependent, multi-year experiments simultaneously, each one’s success partially contingent on the others working out, and the timeline for knowing the real answer on any of them runs to the end of the decade, not the end of this news cycle. The operators worth watching closely aren’t necessarily the ones with the loudest 6G or Open RAN press releases — they’re the ones building genuine, working proof points now, the way Rakuten did with Open RAN, precisely because that’s the only way anyone will know which of these bets actually pays off before 2030 forces the answer on everyone at once.