US operators brush off SpaceX, Anthropic bleeds cash, Amazon gets its spectrum

T-Mobile US CEO Srini Gopalan told investors on his Q3 results call that he sees "no room for an MVNO with SpaceX", closing the door on any resale tie-up with Elon Musk's incoming Starlink Mobile before it has even launched.

His verdict matches Verizon's. CEO Dan Schulman, speaking on his own earnings call, put hard numbers on what a rival terrestrial network would demand: 65,000 to 80,000 macro sites, at a cost of $300 billion to $500 billion over seven or eight years.

"There's no magic wand, there's no secret sauce, there's no shortcut," Schulman said. AT&T chief John Stankey struck the same note a week earlier, questioning the satellite threat and stressing the time and scale needed to match incumbent networks.

Why the market disagrees with the CEOs

The unified front hasn't calmed investors. According to Reuters, AT&T shares fell 5.6%, Verizon lost 1.6% and T-Mobile dropped 5% after reports that SpaceX is seeking access to at least 80,000 US cell sites for Starlink Mobile, with a direct-to-device (D2D) launch targeted for early 2028.

Gopalan did leave one gap in the refusal, noting a resale arrangement would be "a wholly different thing". Whether that distinction survives contact with a 2028 launch remains to be seen.

Bernstein analysts told Reuters the impact itself may still be years away, but warned: "anticipation alone, reinforced by each successive spectrum and infrastructure move, is likely to remain an overhang on sector sentiment."

What does Anthropic's filing actually show?

Away from spectrum, Anthropic filed to list Class A shares on the Nasdaq Global Select Market under the symbol ANT, and its SEC registration statement reads like two different companies stapled together.

The AI developer booked $16.3 billion in revenue in the first half of 2026 — up from $1.2 billion a year earlier and $4.6 billion for the whole of 2025. It also posted a net loss of $80.7 billion. Compute and infrastructure costs accounted for most of its $2.4 billion cost of revenue.

Anthropic itself warns the growth engine may stall, stating in the filing: "We do not expect our revenue will continue to grow at comparable rates, and we cannot assure you that it will continue to grow at all." It attributes the surge to "early-stage adoption dynamics, the nascency of the AI market, and rapid initial customer expansion" — dynamics it says "may not be sustainable over time, and may decline significantly and more rapidly than we anticipate".

Jonathan Tower, founder and managing partner of Catapult, told Fortune that investors risk "confusing the magnitude of the AI opportunity with the profitability of its participants". Extraordinary revenue growth, he noted, doesn't necessarily trend into extraordinary returns.

How big is Amazon's satellite move?

The week's third story connects to the first. The FCC Space Bureau approved Amazon's $11.5 billion acquisition of Globalstar, clearing the path for D2D services on its Leo satellite business.

The approval covers roughly 25MHz of Globalstar's mobile satellite spectrum and authorises access to its C-3 system, a 48-satellite constellation designed to add network capacity. The deal preserves Apple's satellite pact with Globalstar: Amazon has agreed to maintain connectivity for current and future iPhones and Apple Watches.

The FCC said the transaction could improve coverage in underserved areas, strengthen emergency communications and expand industrial and enterprise data services.

Analysts read the deal as a spectrum play dressed as a constellation purchase. FDM CCS Insight principal analyst Luke Pearce called it "a major shortcut in providing satellite services to smartphones" that "strengthens its position against SpaceX's Starlink". Richard Haas, spectrum policy consultant at LS Telecom, put it more bluntly: Amazon is buying Globalstar "for its spectrum assets first and foremost", because companies "aren't willing to wait until regulators start assigning more MSS spectrum".

So the pattern is set: incumbents dismiss the satellite threat verbally, hedge against it financially, and wait for 2028 to settle the argument.