A customer choosing a rival should be an ordinary feature of competition, not an invitation to speculate about its chief executive’s removal. An archived copy of Elon Musk’s response to a post about Delta Air Lines records the exact words: “He will lose his job over this.” The archive dates the reply September 30, 2026. Its target was Delta chief executive Ed Bastian. The accessible evidence establishes the public remark, not an operational plan to make it happen. That distinction separates a disturbing display of power from a demonstrated abuse of it. Delta really is the airline, and Amazon really is the competing supplier. In its March 31, 2026 announcement, Delta said it would begin installing Amazon Leo satellite connectivity on an initial 500 aircraft in 2028. Bastian presented the agreement as a technology-and-cost decision, while the announcement also described deeper integration with Amazon’s cloud services and existing relationships with connectivity providers Viasat and Hughes. This was a future installation programme, not a claim that Amazon’s service was already operating across Delta’s fleet. Nor does the company’s published explanation establish that a personal dislike of Musk determined the procurement decision. The corporate structure makes the dispute more consequential than a billionaire’s bruised ego. SpaceX’s accounts for June 30, 2026 record its acquisition of xAI on February 2, following xAI’s acquisition of X on March 28, 2025. Those accounts place Starlink in the connectivity business and X in the artificial-intelligence business of the same corporate group. The relevant concentration is therefore not merely several companies associated with one celebrity: communications infrastructure and a major public-discussion platform have been brought inside one group. These are corporate assets, however, not a personal bank balance that Musk can simply withdraw. Ownership and control must also be kept separate. Musk’s August 13 ownership filing reports approximately 6.42 billion beneficially owned shares on a basis that includes convertible Class B stock, restricted shares and exercisable options. Its 48.4% figure describes the reported share class under those assumptions, not a straightforward cash stake or percentage of voting power. Class B shares carry ten votes each, against one for Class A. The June offering prospectus projected approximately 82.4% voting control, or 82.3% with the additional-share option fully exercised, and gave Class B holders the right to elect a board majority. Those dated disclosures establish entrenched control; they should not be passed off as a freshly measured October 5 percentage. Operational scale is substantial, but scale is not the same thing as a legally established monopoly. SpaceX reported 12 million Starlink subscription service lines on June 30, 2026, twice the year-earlier figure. Those are service lines, not individual people, and exclude separately negotiated enterprise and government customers. It also reported 77 Falcon launches in the first half of 2026: 60 internal and 17 principally for outside customers. That split matters. The rocket business supplies the satellite business with launch capacity, creating an integrated advantage that a stand-alone internet competitor cannot reproduce merely by buying advertising. It does not, by itself, prove unlawful exclusion. The commercial alternative is real, but switching is not frictionless. United Airlines’ March 2025 installation announcement put the core Starlink installation at approximately eight hours per regional aircraft, expressly excluding removal of existing equipment, testing and aircraft modifications. That is useful evidence from a customer about physical work, not a complete switching-cost estimate. Delta’s published agreement supplies no comparable migration bill. Amazon’s own launch record also documents its use of SpaceX rockets, illustrating how a downstream competitor can simultaneously depend on the incumbent upstream. Such dependence deserves scrutiny, but a launch contract is not evidence that SpaceX has denied or degraded a rival’s access. The threshold for coercive power should be concrete: control over something another party needs, an expensive or impractical exit, and evidence that access is being conditioned on unrelated obedience. In this dispute, a job-loss prediction is not proof of a service cutoff, discriminatory pricing, interference with contracts or manipulation of X’s distribution systems. The crucial investigative evidence would be instructions, contract changes, access logs, procurement communications and testimony from affected parties. Musk’s wording supplies a reason to ask those questions, not permission to assume their answers. Delta’s selection of another supplier is also evidence against the sweeping claim that every customer is already captive. Antitrust offers powerful remedies, but no general offence of being too rich or too influential. The Federal Trade Commission’s explanation of monopolization requires examination of a relevant market, durable market power and improper exclusionary conduct; success through a superior product is lawful. Launch services, airline connectivity, residential broadband and social platforms cannot simply be added together into one convenient monopoly percentage. Federal competition authorities and state attorneys general can investigate within their respective powers. Divestitures can restore competition where an unlawful merger is established, but breaking up a corporate group requires a legal case and a remedy tied to the competitive harm, not punishment for an obnoxious post. Public purchasing provides a less theatrical and potentially more immediate tool. Federal Acquisition Regulation section 6.202 allows agencies, with the required findings, to exclude a supplier from a particular procurement to establish or preserve alternative sources, including for national defence and reliable supply. Government dependence need not become permanent surrender. The Congressional Research Service’s January 14, 2026 account described planned national-security launch contracts involving SpaceX, United Launch Alliance and Blue Origin, including an anticipated value of up to $5.9 billion for SpaceX. That was a multiyear contract figure, not annual government revenue. The policy implication is straightforward: maintaining a second viable supplier may have value even when the first supplier wins on immediate price. Continuity protections deserve equal attention. Legislators could establish narrowly defined service obligations for critical connectivity; purchasers could seek advance termination notice, emergency continuity, documented access decisions, independent dispute resolution and workable migration assistance. Regulators would need appropriate statutory authority, while procurement conditions would need to fit the acquisition and applicable law. The Federal Communications Commission’s January 9, 2026 authorization of further Starlink development demonstrates the licensing relationship, not an unlimited power to impose any desired condition. Interoperability should likewise mean tested technical compatibility, not a political promise that one satellite terminal can automatically use every network. These are proposals for reducing dependence without first dismantling a functioning service. Political-finance reform faces a different legal wall. Federal Election Commission guidance permits unlimited contributions to independent-expenditure-only political committees, commonly called super political action committees. Direct candidate contributions and independent spending are not interchangeable categories. The constitutional barrier became more pronounced on June 30, 2026, when the Supreme Court struck down limits on coordinated political-party expenditures. Disclosure, enforcement against prohibited coordination and voluntary public-financing proposals therefore need to be distinguished from a sweeping spending cap that would confront existing First Amendment doctrine. Anger about concentrated influence cannot substitute for legislation capable of surviving judicial review. Wealth taxation addresses accumulation, not an immediate communications emergency. In their February 2025 study, Tax Policy Center researchers Janet Holtzblatt and Gabriella Garriga examined valuation, avoidance and constitutional obstacles. The Supreme Court’s June 20, 2024 Moore decision expressly left wealth-tax questions unresolved. A tax might reduce fortunes over time, but it would not necessarily remove super-voting control or guarantee uninterrupted service. Conversely, a continuity obligation could protect customers without changing the owner’s net worth. These are complementary policy questions, not interchangeable slogans. Another jurisdiction has already shown both the reach and the resistance of regulation. The European Commission imposed a €120 million transparency fine on X on December 5, 2025 under the Digital Services Act. On September 24, 2026, the United States Justice Department announced an application to support Musk and X’s challenge, arguing that the Commission had overreached in attributing liability across corporate boundaries and to Musk personally. The Commission’s enforcement framework permits substantial fines and corrective measures, subject to judicial challenge. The episode does not prove regulatory capture, but it demonstrates that a powerful company can receive government backing against another jurisdiction’s regulator. It also exposes a central unresolved question: when does common control justify treating nominally separate businesses as one economic unit? (germany.representation.ec.europa.eu) The reporting cutoff is October 5, 2026. Searches found recent material on the Delta dispute, but did not establish a complete October market-share dataset, a comparable five-year series across these different markets, or Delta’s contractual switching costs. The ownership and operating figures above retain their disclosure dates for that reason. The accessible quotation evidence was an archived reproduction rather than a retrievable original X page. Company positions come from public statements and filings; no new interviews or direct company responses were obtained for this article. The evidence supports scrutiny of concentrated control, not a finding that Musk has already used infrastructure to force Bastian’s removal.