Carrier JV, Take Three: Coverage Breakthrough or Familiar Mistake?
Why the AT&T-T-Mobile-Verizon satellite venture looks uncomfortably familiar
May 16, 2026
The U.S. wireless industry has a recurring fantasy. Every decade or so, AT&T, T-Mobile and Verizon convince themselves that if they just stand close enough together, they can will into existence a platform that none of them can build alone. They were wrong in 2010 with ISIS. There are reasons to think they are wrong in 2024 with Aduna. And there are reasons to think they are wrong now with whatever this satellite JV ends up being called.
In the most recent announcement, AT&T, T-Mobile and Verizon agreed to a rare joint venture that aims to make satellite capabilities more widely available to mobile phone customers, with the carriers pooling their satellite partnerships and spectrum resources to better integrate supplementary service into terrestrial mobile networks, is being received in industry trade press as a coverage breakthrough. It is worth pausing on that framing. Is it that the three companies that compete on coverage are agreeing to stop competing on coverage?
The cover story doesn’t survive contact…
The carriers will pool limited spectrum resources to increase capacity, improve the customer experience, and help satellite providers reach more customers through a unified platform. Read that sentence twice. “Pool limited spectrum resources” is what carriers are supposed to bid against each other to acquire. “Help satellite providers reach more customers through a unified platform” means a single chokepoint sitting between every U.S. satellite operator and 95% of the U.S. retail wireless market. This is being marketed as expanded choice. It is the opposite. AST SpaceMobile, Starlink, Globalstar and Iridium today have three separate front doors to the U.S. wireless market and can play them against each other. After this JV, they have one.
The carriers’ own language gives the game away. A technology-neutral innovation platform with industrywide device compatibility… with a standards-based approach is the standard vocabulary deployed when an oligopoly wants to standardize the terms on which it negotiates with suppliers. The satellite operators are the suppliers here. And the throwaway reassurance that existing carrier-satellite agreements will remain in place and the JV partners can continue connectivity efforts independently is not a reassurance. Why announce a JV if the bilateral deals are doing the work?
Softcard is the template.
Memory has been kind to Isis/Softcard. It shouldn’t be. In November 2010, AT&T, T-Mobile, and Verizon officially announced a joint venture known as Isis, which planned to develop a near-field communications-based mobile payments platform. The three carriers announced plans to invest more than $100 million in the project. The Softcard service was discontinued on March 31, 2015, and Google announced that it would acquire certain assets and intellectual property from Softcard. Five years. Nine figures of capital. Endless delays. A name change because nobody at the JV apparently read a newspaper. And the assets were ultimately sold for scrap to the company the carriers had been trying to lock out.
That last point is the one nobody wants to remember. Verizon refused to allow its devices to access Google Wallet because Google Wallet requires access to the “secure element” of a smartphone, even though Softcard has the same requirement. The carriers used network and device control to obstruct a competing platform — not because Softcard was better, but because Softcard was theirs. The market noticed. After these companies blocked Google Wallet for years, the eventual surrender was complete. The joint venture had trouble gaining traction the past five years with several delayed launches, changes in business model, limited merchant and customer participation, layoffs, and one analyst called the rebrand “a new coat of paint on a car that’s not running very well; more is needed to get at the crux of the problem”.
It failed because three companies that compete on retail brand cannot jointly own a retail-facing platform. There is no shared incentive to make it succeed. Each carrier would rather the JV deliver mediocre table-stakes parity than let any of the other two break out. That governance pathology doesn’t go away when you change the product from a payment wallet to a satellite aggregator.
Aduna may be failing in slow motion
The bullish counter is supposed to be Aduna - the global network API JV that was created to combine and sell aggregated network Application Programming Interfaces (APIs) globally, with Ericsson holding a 50% stake and a consortium of twelve major CSPs, including AT&T, Bharti Airtel, Deutsche Telekom, KDDI, Orange, Reliance Jio, Singtel, Telefónica, Telstra, T-Mobile, Verizon, and Vodafone. The pitch is that this is a more sophisticated structure than Softcar - supplier-side, standards-based, hyperscaler-friendly. The pitch deserves scrutiny, and the evidence is starting to accumulate on the wrong side.
Look at what’s actually been said by people who would prefer not to say it. APIs remain a tough sell in the sector. The open source Camara project - the API standard from which Aduna follows - told SDxCentral last month it remains challenging to encourage developer buy-in, with analysts questioning whether telecom operators can reap rewards from the API ecosystem. That is not the language used about a platform that is working ten months in. That is the language used about a platform whose backers are starting to triangulate the exit narrative. Source: SDxCentral
Then look at the architecture. Ericsson is quick to point out that Aduna is not an extension of Vonage’s Global Network Platform (GNP) announced and launched in September of 2022 and has been in development for over two years. In essence, GNP will be subsumed by the Aduna entity and repurposed for its global, industrywide mission. Translation: Ericsson’s expensive Vonage acquisition failed to build a developer business on its own, so it has been repackaged into a JV that distributes the failure across thirteen balance sheets. Note the analyst’s gentle phrasing: Ericsson will likely see the deal as a silver lining in light of its Vonage Holdings investment fallout, which is industry-speak for “this is a vehicle for a write-down.” Source: RCR Wireless NewsSDxCentral
Then look at what the JV is actually delivering. Ten months in, the public narrative is still about adding equity partners and announcing executive hires, not about developer adoption metrics, API call volumes, or revenue. The reason is straightforward: developers don’t want telco APIs. They want cloud APIs. The network APIs are designed to enable network operators to offer services such as service level assurance, fraud prevention and authentication programmatically to application developers, similar to how those same application developers can easily spin up cloud compute instances on cloud providers like Google Cloud and Microsoft Azure — but the operative word in that sentence is “similar.” Developers have those services already, from companies that built developer-first cultures over fifteen years. Telcos did not. A JV does not manufacture one. Source: Comsoc
And the analyst voice the trade press relies on to validate Aduna keeps using conditional tense. “In my view, Aduna Global’s emerging role as a global API exchange has the potential to be catalytic in scaling the network exposure and services market at a multi-regional scope and level, possibly globally”. “Has the potential to be.” “Possibly globally.” Read that without the brand attached and you’d assume the speaker was hedging because the speaker was hedging. Fierce Network
The most plausible path is that Aduna limps on for another two or three years, adds more equity partners as a substitute for adding revenue, gets spun as a long-term bet on 5G monetization, and eventually either gets quietly absorbed back into Ericsson or sold to a hyperscaler at a fraction of invested capital. The shape of the failure may already be visible. The industry has just chosen not to look at it yet.
The satellite JV inherits both failure modes simultaneously
This is the part that should worry us all. The new JV combines the worst structural feature of the past endeavors - three U.S. retail carriers with directly opposed competitive incentives - with the worst structural feature of Aduna - a wholesale aggregation play in a market where the suppliers don’t need an aggregator and the customers don’t want one.
The satellite operators don’t need this JV. AST SpaceMobile already has a direct relationship with AT&T and Verizon. Starlink already has a direct relationship with T-Mobile. Globalstar has Apple. Iridium has its own ecosystem. Every meaningful D2D provider already reaches the U.S. retail market. The only thing the JV adds is a layer of carrier-controlled governance that the satellite operators must now negotiate through, and a “standard” they must conform to. From AST’s or Starlink’s perspective, this is friction, not a feature.
The end customer doesn’t need this JV either. By pooling limited spectrum resources to increase capacity, improve the customer experience, and help satellite providers reach more customers sounds like a consumer benefit until you ask what specifically the consumer gets that they wouldn’t get from the bilateral deals already in place. T-Mobile and Starlink already launched D2D text messaging. AT&T and AST have demonstrated voice and data. The bilateral path was working. The JV doesn’t accelerate it — it reorganizes it. Source: MacTech
The OEMs will route around it. The satellite handoff logic that actually matters happens at the operating system level, on chipsets, in modems. Apple and Google are not going to adopt a U.S.-carrier-defined “common technical specification” that constrains how they design D2D into iOS and Android. They will do what they did to Softcard. They will build their own version, integrate it natively, and the carriers will eventually capitulate.
And the regulators are going to look at this hard. Three carriers comprising essentially the entirety of U.S. retail wireless, jointly controlling spectrum allocation for a new service category, jointly setting technical specifications that suppliers must conform to, jointly negotiating with those suppliers as a unified counterparty. This has the structural fingerprints of a buying cartel. The fact that the marketing language emphasizes underserved communities and reliable connectivity in emergencies is exactly the weak point.
Bottom line
The carriers have run this play three times now. Softcard was a consumer-facing fiasco. Aduna shows early signs of being a supplier-facing one, unfolding more slowly because the metrics are less visible. This new satellite JV combines features of both — three-way carrier governance with a wholesale aggregation thesis.
The base case is roughly two years of glossy press releases, a CEO hire from outside the carriers to signal neutrality, a few demonstration deployments that get extensively covered, the announcement of equity partners as a substitute for the announcement of customers, and a slow drift into irrelevance as Apple, Google, Starlink and AST simply build around it. By 2028 the JV will either have been quietly restructured, absorbed back into one of the parents, or sold for parts. The carriers will then announce a new joint venture. We will write this analysis again.
The kindest thing to say about the satellite JV is that the carriers have learned to make their JV announcements harder to short. The harder question is whether the underlying structural problem, that three retail competitors cannot jointly own a platform, has been solved by any of the three attempts. The evidence so far suggests it has not.