No, SpaceX Isn't Buying Verizon. That's Not Even the Interesting Part.

The real story isn't a Verizon acquisition — it's SpaceX building orbital AI infrastructure at a scale that should worry every telecom operator.

Brian Newman

8/3/20262 min read

A padlock sits on a laptop keyboard with glowing red, green, and blue light trails swirling around, creating a sense of security and cyber awareness.

Verizon has now issued a flat denial: "any assertion that Verizon is being positioned for a sale to SpaceX — or any other entity — is unequivocally false, unsubstantiated, and entirely without merit." That statement exists because current and former Verizon executives were speculating otherwise, and because Elon Musk, when asked directly on the All-In podcast whether he'd buy a carrier, didn't say no. He said, "It's not out of the question. I suppose that may happen."

I spent 35 years inside a major U.S. carrier. I can tell you the denial is probably true and the speculation is still worth taking seriously — just not for the reason most coverage gives it.

SpaceX doesn't need to buy Verizon. Musk's actual playbook, laid out by Light Reading and others, is simpler and more disruptive: don't acquire the operating company, buy the inputs, and build the network from scratch. That means bidding for spectrum at auction and constructing a satellite-terrestrial hybrid network SpaceX owns outright, unencumbered by a legacy workforce, a legacy union contract, or a legacy capital structure. Every acquisition rumor misses this. Musk isn't shopping for Verizon's customer base. He's shopping for Verizon's spectrum rights, and he'd rather win them at auction than pay a control premium for the whole company.

The more interesting story is what SpaceX is actually building, and it isn't a phone company. Following the February 2026 merger with xAI, Grok is the company's anchor tenant for a much larger ambition: orbital AI compute. SpaceX says it wants to launch 100 gigawatts of compute capacity into space annually — roughly a fifth of total U.S. power production in 2025. That number should stop you. It reframes SpaceX not as a satellite internet company angling for a wireless license, but as an AI infrastructure company that happens to already own the rockets, the constellation, and increasingly the compute demand to fill it.

Meanwhile, Verizon is quietly making a parallel bet, just from the terrestrial side. The AI infrastructure race isn't confined to hyperscaler data centers anymore. It's moving into orbit, into fiber, and into every network operator's balance sheet at the same time. Verizon's real strategic question isn't "will Musk buy us." It's whether Verizon can build AI infrastructure value fast enough that being bought — by SpaceX or anyone else — stops being the more attractive path for shareholders.

That's the deal nobody is covering because "SpaceX buys Verizon" is a better headline than "two companies are racing to become AI infrastructure providers using completely different starting assets." But the second story is the one that determines which telecom operators still exist as independent companies in 2030.

Operators evaluating their own AI infrastructure strategy should ask themselves the same question I'd ask any Verizon board member: if a well-capitalized competitor decided tomorrow to build your network from scratch using only spectrum and capital, how much of your enterprise value survives that thought experiment?

Sources: Semafor, Light Reading, PhoneArena

Insights

Empowering businesses with AI and emerging technologies.

Courses

Consulting

Brian@AI-DrivenConsulting.com