FCC Spectrum Rules Weren’t Built for Satellite Mega-Constellations

Last updated 2 weeks ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

SpaceX operates over 9,400 satellites. When Elon Musk announced the company’s acquisition of xAI in a deal valued at $1.25 trillion, the combined entity outlined plans to launch millions more. The stated purpose: building “orbital data centers” to run artificial intelligence models in space, where solar power is constant and cooling is free.

A single company would control more of the radio frequencies that satellites use to communicate with each other than the FCC’s entire regulatory framework was designed to allocate across dozens of competing operators.

SpaceX has authorization to deploy 15,000 satellites for its Gen2 Starlink broadband constellation.

The regulatory system wasn’t built for this.

How Spectrum Allocation Was Designed

Radio frequencies are limited. The same frequency can work for multiple users only if they’re far apart, use it at different times, or coordinate carefully to avoid interfering with each other.

Since its creation in 1934, the FCC’s job has been preventing interference while ensuring no single company controls too much of this public resource. For mobile communications, the agency developed hard limits on how much radio frequency space any company could own. For broadcast television, the FCC restricted how many stations one company could own in a single market.

When satellite communications emerged in the 1960s and 1970s, regulators assumed satellites would follow the same competitive model. A company might own one or two satellites serving specific regions or providing particular services. The regulations built in assumptions about operational size. An operator would deploy a specific number of satellites, with milestone requirements ensuring infrastructure got built.

Most importantly, the regulations assumed companies would stay separate and independent. If Operator A used frequencies 1-10, Operator B would use frequencies 11-20, and Operator C would use frequencies 21-30. As long as companies remained independent with separate financial interests, the competitive structure would hold.

SpaceX’s Incremental Expansion Strategy

On January 9, 2026, the FCC approved 7,500 satellites, bringing the total authorized Gen2 Starlink broadband constellation to 15,000 satellites.

SpaceX never asked for permission to deploy tens of thousands of satellites all at once. It submitted a series of modification requests, each building on the previous authorization. Each modification gets evaluated individually, limiting competitors’ ability to argue that approving it would concentrate orbital resources too heavily in one firm’s hands.

The FCC staff is under pressure to approve applications quickly and clear the backlog. That creates incentives to approve modifications meeting technical standards rather than conduct broader competitive analysis.

The technical standards themselves have shifted in SpaceX’s favor. This represented a deliberate choice: bet that technological advancement and good-faith coordination would prevent catastrophic interference, rather than maintaining strict protection for whoever got there first.

But the new rules removed limits on total interference across all operators. In other words, the rules inadvertently created incentives to deploy as many satellites as possible. A larger constellation can better absorb signal degradation from competitors. SpaceX has the largest constellation.

Vertical Integration: Spectrum Control Plus Computing

If mega-constellations tested the limits of FCC spectrum allocation authority, the SpaceX-xAI merger exposes fundamental flaws in the regulatory framework.

Regulators traditionally worried that one company controlling spectrum would raise prices for internet service. This framework assumes spectrum is mainly for moving data from one place to another.

The merged SpaceX-xAI entity plans to use orbital spectrum differently. It plans to use solar-powered satellites connected by laser beams to handle huge amounts of data with computing capacity to run advanced AI models.

This isn’t about selling satellite internet to consumers. It’s about monetizing computing capacity directly. A customer seeking to train a large AI model would pay significantly more for access to space-based computing than they pay for satellite internet access.

If orbital spectrum becomes primarily valuable for hosting computation rather than transport, then monopoly control over that spectrum translates directly into monopoly control over access to certain types of computing infrastructure.

A competitor would either need to build its own satellites (enormously expensive and takes years) or negotiate with SpaceX-xAI.

SpaceX already controls the launches. Competitors who want to build their own satellite networks have to pay SpaceX to launch them. A competitor seeking to build orbital computing infrastructure would need either to develop its own capability (which requires enormous investment and takes years) or negotiate with SpaceX for launches while trying to develop competitive products that rely on SpaceX infrastructure.

The merger with xAI reinforces this vertical integration, adding an internal customer—AI computing services—that will directly depend on SpaceX’s infrastructure.

SpaceX’s First-Mover Advantage

SpaceX got a head start. Later applicants had to prove their satellites wouldn’t interfere too much with SpaceX’s system and accept stricter limits.

When SpaceX tried to use certain radio frequencies for mobile phone service, the FCC rejected it for technical reasons. But the FCC left the door open for SpaceX to try again in the future.

SpaceX’s advantage gets even bigger because it controls the launches. When competitors like Amazon want to launch their own satellites, they have to negotiate with SpaceX. This cost advantage means SpaceX can build bigger satellite networks for less money.

Amazon has used some SpaceX launches but is also building partnerships with other launch companies to avoid depending on SpaceX. Even Amazon, one of the world’s largest companies, recognizes the risk of depending on a competitor for access to needed infrastructure.

The Merger’s Implications

The $1.25 trillion price tag assumes SpaceX will control space-based AI computing. By connecting SpaceX to AI computing needs, the merger justifies the high price tag and suggests space-based AI computing could solve problems Earth-based systems are hitting.

It’s unclear whether space-based AI computing will actually make money. Analysts say space-based AI computing might not work because chips break down in space, signals take too long to travel, and many AI tasks need instant responses.

Whether regulators need to worry depends entirely on whether Musk’s prediction is correct. If space-based AI computing doesn’t work out, the merger is one company controlling multiple parts of the same business. If space-based AI computing becomes the cheapest way to run AI, and SpaceX controls it, then AI companies would have to negotiate with SpaceX for access. That’s exactly what antitrust law was designed to prevent: one company controlling essential infrastructure and using that power to block competitors.

The national security implications are equally significant. The Pentagon increasingly depends on SpaceX satellites for communications and secret military programs. If space-based AI computing becomes essential and SpaceX controls it, the government becomes even more dependent on SpaceX. If Starlink internet is a national security issue, space-based AI computing would be an even bigger one.

Regulatory Options

The FCC has significant authority on paper. The Communications Act says the FCC must approve any license transfer if it’s in the public interest. The FCC could theoretically use this rule to block the merger because of competition concerns.

But the FCC usually lets the Justice Department and FTC handle merger reviews instead of blocking them itself. The Trump administration is unlikely to block mergers in important industries.

The Justice Department and FTC review large business deals under federal law. The $1.25 trillion valuation almost certainly exceeds reporting thresholds, triggering federal review. The 2023 merger guidelines say a merger can be illegal if it lets one company block competitors from getting what they need or control too many parts of the business.

To block the merger, regulators would need to prove SpaceX-xAI could and would block competitors from getting what they need—which requires showing there’s no good alternative.

Congress could fix this problem by passing new laws. A bill by Senators Cruz and Welch, the “Satellite and Telecommunications Streamlining Act,” would speed up how the FCC approves satellite licenses. But the bill doesn’t address whether one company is getting too much power.

A better approach would set clear rules about how much satellite spectrum one company can control.

The alternative is to let other companies like Amazon and OneWeb compete with SpaceX. Amazon’s satellites are smaller and it’s behind schedule. By the time competitors catch up, SpaceX-xAI might be so far ahead that nobody else can compete.

The FCC’s Modernization Effort

Recognizing its rules don’t work for mega-constellations, the FCC is rewriting them. In October 2025, the FCC proposed new rules for how to license satellites. The new rules would let companies build satellites faster if they meet basic safety standards.

But the modernization reveals a real problem: the FCC is trying to speed up approvals when the real issue is deciding who should get spectrum. Even if the FCC approves applications faster, the real question is: how should it decide who gets to use radio frequencies when everyone wants them?

If dozens of companies all want to launch thousands of satellites, but space can’t handle all of them, how should the FCC decide?

The current approach avoids making this choice. The FCC sets rules about how much interference is allowed and requires newer companies to not interfere too much with earlier ones. But it doesn’t set a total limit on how many satellites can operate or choose between competing companies.

This means the FCC keeps approving applications until space gets too crowded to work. At that point, the FCC will either start rejecting applications or allow even more interference.

This reflects a shift from carefully controlling spectrum to letting companies build whatever they want. Under the old approach, the FCC would carefully decide who gets to use radio frequencies. Under the new approach, it sets minimum rules and lets companies launch satellites, hoping they’ll figure out problems. That works when the problems are engineering issues. It doesn’t work when one company controls so much that it becomes a monopoly.

International Coordination Challenges

The challenge is further complicated by the international dimension of satellite spectrum allocation. An international organization called the ITU decides how countries share radio frequencies. Operators seeking global service must tell the ITU and other countries about their plans and resolve disagreements.

The ITU works differently than the FCC in important ways. The ITU gives priority to whoever applies first for radio frequencies. This encourages companies to apply early, even before they’ve finished designing their systems.

SpaceX has had mixed experiences with international coordination. In October 2023 and March 2024, the ITU sided with Iran against SpaceX’s claim that it shouldn’t have to track where Starlink users are. The ITU said countries can require SpaceX to verify where users are for legal and security reasons.

The FCC and ITU don’t always agree. The FCC lets newer companies accept interference in ways the ITU doesn’t allow. This creates a problem: American companies can operate in ways that violate international rules. Many countries, especially in Europe and China, worry about too many satellites and want stricter rules.

What Happens Next

Federal regulators will soon decide how to handle the merger and regulate mega-constellations. One approach accepts that one company will dominate and tries to prevent unfair practices through rules about who can own what and who gets access. Another approach tries to block the merger or use antitrust law to break up the company. A third approach keeps current rules and hopes competition prevents one company from taking over.

This decision will determine who controls the satellite infrastructure governments depend on. The FCC’s rules were designed for dozens of companies competing for limited radio frequencies. One company now wants to launch millions of satellites. The old rules don’t know how to handle one company wanting more than the entire system was designed for.

This isn’t a technical problem. It’s a failure to make rules, and it will decide who controls space-based computing for decades.