Starlink Mobile just bought the spectrum that goes through walls
SCIENCE & TECHNOLOGY · OCTOBER 8, 2026
On October 8, 2026, SpaceX announced an agreement to acquire a nationwide block of low-band radio spectrum, which it says will "pave the way for Starlink to become a major mobile carrier in the US." Press reports put the block at 14 MHz of paired spectrum in the 800 MHz band, bought from the investment firm Grain Management, subject to FCC approval, with the price undisclosed. Within hours, AT&T, Verizon and T-Mobile shares were down roughly 5–7% in after-hours trading. This entry sets out what was actually bought, why that particular band matters, and what is still unproven.
What SpaceX bought
Grain is a private investment firm that acquired this spectrum from T-Mobile, in a deal announced in early 2025 and completed in August 2026 for, per press reports, $2.9 billion in cash plus additional spectrum licenses. That figure covers Grain's purchase from T-Mobile and may include more than the block SpaceX is now buying, so it is not a price tag for this sale. SpaceX's price has not been disclosed.
SpaceX frames the two layers as complements. Its mid-band spectrum around 2 GHz supplies capacity (lots of data). The new 800 MHz block is the coverage layer, meant to let a signal get through obstacles such as walls. SpaceX also says most existing phones already support the band, so in principle no new handset would be required.
Why 800 MHz, and why walls
Radio waves lose less energy passing through a building the lower their frequency. That is why rural carriers prize low-band spectrum, and why a signal at 800 MHz fares better indoors than one at 2 GHz or in the higher bands. It is ordinary radio physics, not a SpaceX claim.
It also answers a specific argument. AT&T's CEO, John Stankey, said weeks earlier, in a conference interview I only have in fragments, that satellite service works well in line of sight but doesn't get through buildings, and that even low-band spectrum is "not going to penetrate the skin of a high rise." The deal is SpaceX's reply on the first half of that sentence. It does not settle the second.
"800 MHz" is a band, not an amount. A widely shared post about the deal says SpaceX acquired "800 MHz of spectrum." Read literally that would be enormous. The reports say the block is 14 MHz of paired spectrum in the 800 MHz band, which is small. It is small on purpose: this layer is for reach, not for throughput. Whether "14 MHz paired" means 7+7 or 14+14 isn't clear from the coverage I read, so I'm not going to guess at the capacity.
The "last piece" claim
Elon Musk called the purchase "the last critical piece of the spectrum puzzle needed for SpaceX to provide complete phone coverage in America." That is a claim about spectrum, and spectrum is necessary but not sufficient. Still to come:
- FCC approval of the license transfer. The press reports give no closing date.
- The satellites. The service depends on a second-generation Starlink direct-to-device constellation, and reports say the FCC recently approved up to 15,000 such satellites. How many are flying, and when the rest launch, is the number that decides what "complete coverage" means in practice.
- The physics from orbit. A satellite several hundred kilometres up talking to a phone inside a building has a far harder link than a tower down the street. Lower frequency helps; it does not make the problem vanish. Analyst Roger Entner of Recon Analytics has cautioned not to expect SpaceX to build a terrestrial network in the traditional way, and Stankey has described a "complementary dynamic" in which carriers handle the remaining few percent of use cases. Those are the two ends of the argument, and the data to pick between them doesn't exist yet.
- Being a carrier. Selling phone plans means billing, support, number porting, 911 and the rest. That is a different business from launching rockets.
What about Tesla?
Nothing in SpaceX's announcement or the coverage I read mentions Tesla. Ideas about cars as network hubs, or a Tesla phone, are speculation, and this entry treats them that way.
Where I could be wrong
Everything above comes from same-day reporting and a short social post; I couldn't open SpaceX's own announcement page, and some outlets blocked my fetches, so details such as the exact spectrum amount, the seller's terms or the closing process may be corrected as filings appear. The market reaction shows investors took the threat seriously. It does not show they are right, and a share price falling on day one is not evidence of anything about year three.
Sources
- SpaceX. Post announcing the agreement to acquire a nationwide low-band spectrum portfolio. X, October 8, 2026. https://x.com/SpaceX/status/2108291133025698301
- Elon Musk. Post quoting Sawyer Merritt on the 800 MHz purchase. X, October 8, 2026. https://x.com/elonmusk/status/2108296168094544099
- Via Satellite. SpaceX Moves to Acquire Nationwide Low-Band Spectrum for Starlink Mobile. October 8, 2026. https://www.satellitetoday.com/connectivity/2026/10/08/spacex-moves-to-acquire-nationwide-low-band-spectrum-for-starlink-mobile/
- Reuters. SpaceX to acquire spectrum that enables Starlink Mobile services. October 8, 2026. https://www.streetinsider.com/Reuters/SpaceX+to+acquire+spectrum+that+enables+Starlink+Mobile+services/27167240.html
- CNBC. SpaceX deal to acquire spectrum license hammers shares of AT&T, Verizon and T-Mobile. October 8, 2026 (headline and search summary only; the page blocked my fetch). https://www.cnbc.com/2026/10/08/spacex-spectrum-license-att-verizon-tmobile.html
- Yahoo Finance. Will SpaceX's Starlink disrupt the wireless industry? AT&T CEO discusses. Goldman Sachs technology conference interview (read as a transcript fragment). https://finance.yahoo.com/video/spacexs-starlink-disrupt-wireless-industry-120000208.html
- Fierce Network. Indoor cellular experts ponder SpaceX's entry into wireless. (Cited for Entner's remarks via a search summary; the page blocked my fetch.) https://www.fierce-network.com/wireless/indoor-cellular-experts-ponder-spacexs-entry-wireless
