SpaceX’s push into the US mobile market, not its home satellite internet business, is what sent AT&T’s shares lower in early October, according to an analysis published on 11 October by the financial site Ainvest. The carrier’s stock fell after SpaceX agreed to buy low-band spectrum from an investment firm and after SpaceX President Gwynne Shotwell told the company’s first earnings call since its stock-market listing this year that she expected to win customers away from traditional wireless carriers.
That wireless-versus-fiber thesis belongs to one source. Ainvest says the piece was written by an AI research-and-writing agent and publishes it with a disclaimer that it is information only, not investment advice. No other analyst or carrier quoted below draws the same distinction, so the attributions in this report separate Ainvest’s reading from AT&T’s own disclosures and from reporting by PCMag.
Key facts
- SpaceX agreed to buy up to fourteen megahertz of 800 MHz airwaves from Grain Management, an investment firm, in a deal reported at around $8 billion, Ainvest reported.
- AT&T’s stock fell about seven percent in an afternoon, which Ainvest attributes to investors reading the spectrum deal together with Shotwell’s comment that Starlink would take “quite a few” customers from carriers.
- The US Federal Communications Commission approved a next-generation Starlink Mobile constellation of 15,000 satellites with more than one hundred times the bandwidth of the current one, per Ainvest.
- AT&T chief executive John Stankey told Axios that fiber is three times faster than satellite and that the plan is not economically or technically feasible, according to Ainvest.
What SpaceX says
Shotwell’s earnings-call line about taking “quite a few” carrier customers is the only on-record SpaceX comment in the three reports, and it was made before Stankey’s feasibility objection rather than in answer to it. Neither Ainvest nor PCMag records any SpaceX response to AT&T or to T-Mobile, and the company has not published a consumer launch date, a coverage map or a retail price for a Starlink-branded wireless plan.
Why it matters
Ainvest’s argument is that the two sides are measuring different things. Starlink’s home service, it says, was never aimed at fiber: it sells into places fiber does not reach. The spectrum buy, by contrast, reaches for AT&T’s wireless postpaid customers, which Ainvest calls the single biggest driver of the company’s revenue and its steadiest cash — the business supporting a dividend yield that Ainvest data puts above six percent.
One figure in that case comes from AT&T rather than from the analysis: the company disclosed that at the end of the second quarter 40.5 percent of AT&T Fiber customers also subscribed to AT&T wireless, a number Ainvest reads as evidence the carrier is using fiber to defend the wireless book.
Ainvest notes that low-band 800 MHz travels farther and penetrates walls better than satellite frequencies, which is why it reads SpaceX’s purchase of 800 MHz spectrum as a licence to enter wireless rather than a service that has taken anyone yet.
What AT&T says
AT&T has not commented publicly on the Grain Management purchase itself. Its most recent on-record position dates to July, and was made about satellite-to-phone service in general.
In an interview with Axios, Stankey said of fiber, “Nothing beats it,” as quoted by Ainvest. In July, on AT&T’s second-quarter earnings call and in a CNBC interview reported by PCMag and carried by Yahoo Finance, he said new competitors were coming to the game very late and would have to play catch-up against decades of infrastructure investment.
His technical objection, per PCMag, is coverage quality rather than headline speed: it is unclear whether space-based 5G can penetrate buildings made of concrete and metal, and customers now expect service to work inside hospitals, on university campuses, in stadiums and in tall buildings. Stankey also argued that once fiber is in the ground, there is no customer AT&T should not win, citing symmetrical gigabit and multi-gigabit service and the lowest marginal operating cost.
Rival carriers have been blunter. Ainvest quotes T-Mobile’s chief financial officer saying that no number of satellites or amount of spectrum can break the fundamental laws of physics; T-Mobile executives have made the same argument elsewhere.
The obstacles Ainvest identifies
Ainvest accepts that converting an authorisation into a network is hard. It says regulators do not let a homeowner radiate cellular signals without permission, and puts the cost of building enough terrestrial sites to match a conventional network at a hundred billion dollars or more — a figure it credits to analysts it does not name, and which should be read as the site’s own estimate rather than a sourced number.
A spectrum authorisation, it writes, is a ceiling; paying subscribers are the fact on the ground, and the FCC cleared 15,000 more Starlink Mobile satellites without any service yet taking a customer.
The test Ainvest proposes is where Starlink Mobile actually wins users. If it only serves dead zones and remote corners the incumbents do not cover, it remains a complement. If it starts taking price-sensitive subscribers in areas AT&T already covers well, Ainvest says, that is a real injury to the cash flow behind the dividend.
Price and availability
Neither the spectrum deal nor the FCC decision changes what anyone pays today. Starlink’s satellite internet home service remains the only Starlink product with published pricing, and the next-generation constellation is licensed for the United States, where Ainvest reads the 800 MHz holding as the step that would let SpaceX sell wireless under its own name instead of through a carrier partner. For readers in Zimbabwe and elsewhere in Africa, no Starlink Mobile tariff, licence or launch window has been announced in any of the three reports; the home satellite internet subscription is what is on sale.
Background: how AT&T is hedging
AT&T is expanding fiber through a joint venture with the infrastructure investors Global Infrastructure Partners and CPP Investments, pushing its network toward sixty million locations by 2030, Ainvest reports.
The carrier has also backed a competitor to SpaceX’s satellite-to-phone plans. According to PCMag, AT&T has invested in AST SpaceMobile, a Texas-based company working to launch its own version of Starlink Mobile, which PCMag said in July was expected sometime the following year.
Frequently Asked Questions
How fast is Starlink Mobile supposed to be?
The next-generation constellation behind Starlink Mobile promises speeds of up to 150Mbps for 5G satellite connectivity direct to phones, according to PCMag’s July report carried by Yahoo Finance. The same report said SpaceX is separately working on a gigabit upgrade for its home service.
How many Starlink subscribers are there?
Ainvest put Starlink at twelve million subscribers across rural homes, remote sites and boats, and said the home product grew for years without disturbing the big US carriers because it sells where fiber does not reach.
Would AT&T sign a wholesale deal with Starlink for dead zones?
No. John Stankey told CNBC in July that he does not need a wholesale arrangement to reach the 2% of traffic he cannot get at today, because AT&T already reaches 98%, according to PCMag. He said he would think differently only if those numbers were reversed.
How fast is AT&T’s fiber business growing?
Stankey said on AT&T’s July earnings call that 2026 would be its largest year ever for fiber expansion, with plans to reach 8 million new locations, including 4 million acquired from Lumen, as reported by PCMag. In the second quarter AT&T added 367,000 fiber customers and 279,000 fixed wireless users.
Follow Starlink News on Google. Make us a preferred source to see more of our reporting in Google search results.
![Using Your Own Router with Starlink: Bypass Mode [2026]](https://starlink-news.com/wp-content/uploads/2026/09/using-your-own-router-with-starlink-bypass-mode-20-f6c09f95-120x86.png)
