Elon Musk Is Not Just the World's Richest Man — He Is the Landlord of the AI Era

SpaceX's world record IPO formalised the most consequential infrastructure concentration in the history of technology.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://open.substack.com/pub/johanosteyn/p/elon-musk-is-not-just-the-worlds?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

On 12 June 2026, SpaceX began trading on the Nasdaq under the ticker SPCX at 135 dollars per share, valuing the company at 1.77 trillion dollars — the largest listing price in history. It raised 75 billion dollars. Institutional demand exceeded available shares by four times. Saudi Arabia’s Public Investment Fund and the Kuwait Investment Authority each booked between 1 and 5 billion dollars of stock. Retail investors placed orders exceeding 100 billion dollars. Elon Musk, who controls the company through a dual-class share structure in which his Class B shares carry ten votes per share to the single vote of Class A shares — giving him the power to elect the board and control all matters requiring shareholder approval — was worth approximately 982.6 billion dollars at the IPO price, closing in on the first trillion-dollar individual fortune in history.

The financial story is extraordinary. It is not the most important story. The most important story is what Elon Musk now controls — and what no governance framework was designed to govern.

CONTEXT AND BACKGROUND

The Colossus supercomputer cluster — one of the largest AI computing facilities in the world — sits on SpaceX’s balance sheet. In May 2026, SpaceX entered into cloud services agreements with Anthropic covering access to compute capacity across Colossus and Colossus II at 1.25 billion dollars per month through May 2029, with either party able to terminate upon 90 days’ notice. Days after the S-1 was filed, Musk posted on X that the deal was actually a 180-day lease with 90-day mutual cancellation — a characterisation the prospectus does not support and that legal experts described as confusing to investors. Whether the agreement runs to May 2029 or ends in six months is not a technical footnote. It is the governance question at the heart of what Anthropic’s infrastructure dependency actually means in practice. The AI company most publicly committed to safety operates its core compute infrastructure under terms that its own landlord cannot consistently describe.

New Street Research initiated coverage of SpaceX ahead of its listing with a price target of 165 dollars per share — implying 22 per cent upside from the IPO price — arguing that investors were systematically underestimating the long-term value of Starlink, artificial intelligence, and future space infrastructure. The firm’s sum-of-the-parts analysis valued Starlink’s broadband business at a scale capable of supporting over 280 million subscribers and more than 100 billion dollars in annual revenue. Its AI and space-based computing segments were assigned a combined value of over 1.5 trillion dollars, with xAI alone valued at 575 billion dollars, integrated infrastructure at 325 billion dollars, and orbital data centres at 650 billion dollars — reflecting the firm’s view that owning the underlying physical stack warrants a substantial premium over rivals who merely build models. Third-party launch operations, despite their historical significance, were characterised as only a small moving part in this picture.

To that infrastructure portfolio, add political access at the highest levels of the US government, where Musk has served in a senior advisory capacity at the Department of Government Efficiency. The same government that this week issued an export control directive pulling Anthropic’s models offline — the models running on infrastructure Musk’s company controls — is one in which Musk has unprecedented individual access and influence.

INSIGHT AND ANALYSIS

The governance question that the SpaceX IPO makes unavoidable is not about Elon Musk personally. It is about what it means for the global AI ecosystem when its critical infrastructure layer — compute, connectivity, launch capability, and orbital data infrastructure — is concentrated to this degree under a single voting control structure, subject to the commercial, political, and personal decisions of one individual.

The historical precedents for this kind of infrastructure concentration are not reassuring. Ukraine’s experience with Starlink makes the dependency dynamic concrete and documented. Musk has stated publicly that Starlink is the backbone of the Ukrainian army and that the entire front line would collapse if he turned it off. In 2022, he reportedly ordered a shutdown of Starlink coverage during Ukraine’s Kherson counteroffensive, directly disrupting drone operations and artillery coordination. Reports have also indicated that the United States used Ukraine’s dependency on Starlink as leverage in broader negotiations. Foreign Policy has framed the conclusion precisely: when a private supplier can decide which operations a front-line state is permitted to conduct based on personal intuition, the relationship has ceased to be commercial — and the same logic applies to the AI infrastructure layer that organisations across the world are now building their operations on.

Morgan Lewis, in a May 2026 analysis of what it calls the new AI corridor, sharpens the strategic argument further: for decades, global infrastructure strategy revolved around oil pipelines and shipping lanes, but the race for AI leadership now resembles the energy geopolitics of prior decades, with the next decade of AI development determined not only by those who build the best models, but by those who control the underlying digital infrastructure. SpaceX, after its IPO and its xAI acquisition, sits at the intersection of every one of those chokepoints simultaneously. The analogy to energy markets is not rhetorical. It is structural — and the SpaceX IPO is the moment at which that concentration became permanent and publicly visible in a way that boards, governments, and regulators can no longer overlook.

IMPLICATIONS

For boards and executives, the SpaceX listing raises a question that is more operational than philosophical. If the infrastructure your organisation’s AI systems depend on — directly or through the AI providers you use — is ultimately controlled through the voting power of a single individual whose commercial incentives, political relationships, and personal decisions can affect its availability and terms, how does that risk appear in your governance framework? For most organisations, the answer is that it does not appear at all. AI infrastructure dependency is modelled as a vendor risk, not as a concentration risk with geopolitical and personal dimensions.

That modelling is now inadequate. This week provided two concrete demonstrations of why. The Commerce Department’s export control directive pulled Anthropic’s models offline for every customer in every country without notice, without consultation, and without appeal. The dependency chain that produced that outcome runs from a US government official to an AI company to a supercomputer cluster controlled by a man who is also an informal senior adviser to the government that issued the directive — and accessible to Anthropic under an agreement either party can exit with 90 days’ notice. For organisations in South Africa and across Africa that are building AI strategies on US-hosted infrastructure, the SpaceX IPO is not a financial event. It is a risk disclosure.

Anchor analyst James Bennett (ITWeb article) offers the most practically useful observation for organisations considering exposure to the AI infrastructure layer: these companies have listed so late that most of the economic value has already been preserved for private shareholders, and for everyday investors there are likely far better opportunities in the broader AI ecosystem. That caution carries additional weight when read alongside the financial picture SpaceX disclosed ahead of its listing. Despite revenue of 18.6 billion dollars in 2025, the company posted a net loss of 4.9 billion dollars for the year, widening through the fourth quarter because of heavy spending on rockets and AI infrastructure — a cash-burning trajectory that continued into 2026. Gulf sovereign wealth funds and institutional investors have already captured the primary value creation. For boards and executives deciding whether AI infrastructure exposure belongs on their balance sheet, the history of stocks going public at record valuations while posting losses at this scale is not an argument for urgency.

The more durable strategic question is not whether SpaceX stock will outperform the broader market. It is whether the concentration of AI infrastructure that SpaceX’s IPO formalises will be subject to the kind of regulatory scrutiny, antitrust examination, and democratic governance that infrastructure of this strategic importance has historically required — and whether the institutions responsible for that oversight are moving at the speed the concentration demands.

CLOSING TAKEAWAY

The world has had infrastructure monopolies before — railroads, telecommunications, energy utilities. In every case, the moment at which the concentration became visible was also the moment at which the governance response became necessary. In every case, the governance response arrived later than the concentration did, and the costs of that latency were borne by the people and institutions most dependent on the infrastructure and least able to influence its terms.

The SpaceX IPO is that moment for the AI era’s infrastructure layer. Elon Musk controls the voting power of a company that owns a significant portion of the infrastructure the AI era runs on, leased to the companies that depend on it under terms that can be unwound in a quarter. The governance question that raises is not about him. It is about what every board, every government, and every organisation that depends on that infrastructure is doing to ensure that their dependency is understood, managed, and not simply assumed to be permanent.

Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net