Elon Musk’s SpaceX has filed for what could be the largest initial public offering in history, according to NPR reporting on 20 May 2026. The filing, which Al Jazeera also confirmed, reveals the company’s blockbuster spending on both rockets and artificial intelligence. If the offering proceeds at expected valuations, it would make Musk substantially wealthier than he already is — an outcome that arrives at a moment when the billionaire retains significant influence over federal contracting decisions through his role advising the Trump administration’s Department of Government Efficiency. The filing comes in the same week that Musk’s other AI venture, OpenAI competitor xAI, was in the news, and as Meta announced the elimination of 8,000 jobs in a separate but related demonstration of how artificial intelligence is reshaping labour markets across the technology sector.
The received wisdom
The bullish case on SpaceX is compelling, and it has been made by serious people who are not reflexive Musk enthusiasts. The company has achieved things that government space agencies spent decades failing to accomplish: reusable orbital rockets, drastically reduced launch costs, and a satellite-broadband network — Starlink — that has had measurable military utility in Ukraine and commercial applications worldwide. SpaceX’s cost-per-kilogram to orbit has fallen by an order of magnitude compared to the Space Shuttle era. This is not hype; it is engineering. An IPO would allow broader public participation in a genuine wealth-creation story and would subject the company to the scrutiny of public markets. The progressive critique that Musk is simply a subsidy farmer misses the genuine technical innovation and ignores the counterfactual: would Boeing or Northrop Grumman have produced reusable rockets faster on similar contract terms? The evidence is not flattering.
A different read
All of that is true — and it is also insufficient. The SpaceX story is one of those cases where the achievement is real and the structural problem is equally real, and pretending one cancels the other out is a form of intellectual convenience.
The core tension is this: SpaceX is among the largest recipients of federal government contracts in the aerospace and defence sector. NPR’s reporting notes the company’s massive spending on both rockets and AI — spending that is significantly enabled by NASA contracts, Department of Defence launch contracts, and regulatory decisions made by agencies that Musk has, simultaneously, been in a position to influence. The head of DOGE, which has been auditing and restructuring federal agencies, is also the primary beneficiary of a public offering that will be partially capitalised by federal dollars. This is not a hypothetical conflict of interest. It is an actual one.
Conservatives who are rightly suspicious of revolving-door corruption in the defence-industrial complex — the lockheed-Boeing-Pentagon triangle that Eisenhower warned about in 1961 — should apply the same scepticism here. The fact that the product is impressive does not dissolve the governance problem. The historical analogy is instructive: Andrew Carnegie’s steel was also impressive, and the Gilded Age’s infrastructure achievements were real. What the Progressive Era eventually concluded was that allowing a single private actor to simultaneously provide essential public infrastructure and extract public subsidy — while shaping the regulatory environment — was not compatible with republican self-government, regardless of how good the steel was.
The IPO also crystallises a question about the AI economy more broadly. The same week that SpaceX files for a record listing, Meta is eliminating 8,000 jobs while telling shareholders it is pivoting aggressively to artificial intelligence. Standard Chartered, according to BBC Business, is cutting thousands of roles for the same reason. The productivity gains from AI are real and measurable. The question that neither the utopian tech-optimists nor the Luddite critics have satisfactorily answered is: who captures them? In the current environment, the answer appears to be: the equity-holders of the companies deploying AI, and the government-adjacent entrepreneurs whose business models depend on state infrastructure and contracts. The workers whose labour is being replaced will be offered, at best, retraining programmes.
SpaceX’s IPO should proceed, subject to the normal mechanisms of market disclosure. But those mechanisms are only legitimate if the underlying conditions are fair — and a company whose CEO advises on federal contracting while collecting billions in federal contracts is not operating in a fair competitive environment by any normal definition of that term.
What to watch
- The IPO prospectus disclosures: specifically, the percentage of revenue derived from federal contracts versus commercial customers, and the terms of any DOGE-adjacent regulatory relief or contract modifications made during Musk’s tenure in the administration.
- Congressional appetite for conflict-of-interest legislation: some bipartisan interest has been expressed but no bill has moved in committee.
- Whether the IPO valuations hold in secondary trading — a test of whether SpaceX’s commercial revenues can eventually replace the government subsidy floor.
- The pace of Meta and similar companies’ AI-driven layoffs, and whether the political class begins framing this as a policy problem rather than a market outcome.
— J