GPT-5: These billionaires’ interconnected gains illustrate how AI infrastructure has evolved into a self-reinforcing oligopoly spanning hardware, cloud, and capital markets.

“Circular Wealth and the AI Infrastructure Oligopoly: Where Prosperity Meets Antitrust”

by ChatGPT-5

The Forbes report and accompanying table reveal a striking concentration of wealth and power around the global AI infrastructure boom. Over just nine months, 20 billionaires added $460 billion to their combined fortunes — with Larry Ellison alone gaining $139.7 billion. The data shows not only the astronomical speed of capital accumulation but also the structural entanglement of the companies behind it: Oracle, Nvidia, Google, Meta, Microsoft, AMD, and emerging hyperscale players such as CoreWeave. These firms are not simply competitors — they are co-investors, customers, and financiers of one another’s infrastructure, forming an increasingly interdependent ecosystem that blurs traditional market boundaries.

1. Circular Deals and Antitrust Red Flags

The deals described in the report — Nvidia investing up to $100 billion in OpenAI so OpenAI can purchase Nvidia GPUs, and AMD granting OpenAI stock warrants tied to GPU usage — exemplify what economists call circular value loops. Such arrangements, while efficient for capital-intensive infrastructure build-outs, create substantial antitrust and competition law exposure in several jurisdictions.

United States.
Under the Sherman and Clayton Acts, regulators could view these cross-financing and reciprocal-supply agreements as mechanisms that reduce competition by vertically integrating critical components of the AI supply chain — from compute hardware (Nvidia, AMD) to cloud platforms (Oracle, Microsoft) and AI models (OpenAI, Anthropic). The Federal Trade Commission (FTC) and Department of Justice (DOJ) have already expressed concerns about “platform entanglement” and self-preferencing in AI ecosystems. A scenario where the same companies fund, supply, and profit from each other’s capacity expansion could constitute de facto collusion, particularly if competitors are excluded from access to compute resources at fair market rates.

European Union.
The European Commission’s Digital Markets Act (DMA) and Competition Law (Articles 101 and 102 TFEU) give Brussels even sharper tools. The European Commission has already opened preliminary inquiries into AI compute supply, cloud pricing, and exclusivity clauses. Nvidia, Oracle, and Google could be targeted under Article 102 for “abuse of dominance” if their reciprocal arrangements foreclose smaller AI model developers or national data centers. The EU is also likely to scrutinize CoreWeave’s multi-year exclusive contracts with OpenAI, Meta, and Microsoft, which could breach the EU’s vertical-restraint rules if they restrict switching or interoperability.

United Kingdom.
The UK’s Competition and Markets Authority (CMA) has positioned itself as a global leader in AI oversight. Its AI Foundation Model Review already signals concern about concentration in compute, data, and cloud. Deals like Nvidia-OpenAI or AMD-OpenAI, which intertwine supply and equity, would fall squarely within the CMA’s definition of “ecosystem entrenchment.” The UK regulator could impose behavioral remedies — such as transparency obligations or data-sharing requirements — similar to those levied against cloud providers in the past.

Asia.
In Japan, the Fair Trade Commission has shown growing willingness to police digital giants, and in South Korea, the KFTC’s scrutiny of vertical integration (e.g., in semiconductor and cloud sectors) could extend to AI infrastructure. China presents a paradox: while state-aligned giants are building similar infrastructure loops domestically, the Chinese government has imposed antitrust penalties on Alibaba and Meituan to signal that even national champions must submit to state-defined “fair competition.”

2. The Emerging “AI Infrastructure Cartel”

The Forbes data underscores that the same small constellation of firms control compute chips, cloud capacity, and the AI models that depend on them. Oracle’s $18 billion debt issuance, Nvidia’s daily stock sales, and CoreWeave’s $29 billion leveraged expansion reflect not just growth but systemic risk concentration. When the top five players are simultaneously supplier, customer, investor, and beneficiary of each other’s contracts, market forces are replaced by self-referential feedback loops — a hallmark of oligopolistic coordination.

Antitrust law historically reacts to such dynamics only after damage accumulates. The late-1990s Microsoft case and the EU’s 2018 Google Shopping decision show that regulators intervene once vertical foreclosure becomes undeniable. The AI infrastructure market, however, is global, capital-intensive, and politically charged: data centers are framed as national security assets, blurring the line between market regulation and industrial policy.

3. Economic Contributions and Political Immunity

The question, then, is whether these billionaires are likely to actually get into trouble. On one hand, their intertwined wealth raises legitimate competition concerns; on the other, their projects create thousands of high-skill jobs, stabilize energy demand in some regions, and attract government subsidies. Oracle and Microsoft are building massive data centers in the U.S., the UK, and the EU, often in partnership with local authorities who view them as anchors of “digital sovereignty.” Nvidia’s AI chip production fuels entire secondary economies — from Taiwan’s semiconductor sector to U.S. clean-energy projects tied to data-center cooling.

Politically, these contributions create immunity through indispensability. Governments pursuing AI competitiveness are reluctant to hinder the very firms that enable it. U.S. national-security arguments (e.g., reducing dependence on Chinese compute) give Nvidia and its allies a quasi-strategic status. In the EU, policymakers face a dilemma: they need antitrust enforcement to protect smaller AI entrants but also depend on American chip and cloud giants to meet AI industrial-policy targets. This tension often results in delayed or symbolic enforcement — heavy on rhetoric, light on disruption.

4. Philosophical Reflection: Can Billions Defy Law?

The final question transcends economics: Do billions nullify regulation?

In practice, extreme wealth tends not to abolish laws but to bend the timeline and narrative of their enforcement. Financial power buys patience, influence, and framing control. A billionaire can fund lobbying, delay proceedings, or recast antitrust scrutiny as a threat to innovation and national progress. Over time, regulators may internalize the argument that curbing such actors equals undermining economic vitality.

Yet there is an irony in this logic: the very concentration of wealth that accelerates innovation also undermines the legitimacy of the systems meant to govern it. If the same handful of individuals can reshape global infrastructure and extract $460 billion in a year through self-referential deals, regulatory inaction ceases to be neutrality — it becomes complicity.

The philosopher Hannah Arendt warned that when political institutions become dependent on private actors for their own survival, power ceases to be shared and becomes administered. Today’s AI billionaires may not seek to overthrow states, but their infrastructure has become the substrate on which economies, governments, and even democratic discourse now run. In such a world, regulation is not nullified by billions — it is absorbed by them.

In conclusion:
These billionaires’ interconnected gains illustrate how AI infrastructure has evolved into a self-reinforcing oligopoly spanning hardware, cloud, and capital markets. While U.S., EU, and UK regulators have the tools to act, the economic gravity and political utility of these firms make serious enforcement unlikely in the near term. The deeper philosophical challenge is not whether antitrust law can catch up, but whether societies can maintain the moral and institutional distance necessary to govern entities on which their digital existence now depends.

Grok: A Plea for Anti-Billionaire Regulation: Safeguarding Society, Civil Rights, and Democracy. The benefits of such regulation are profound.

PASCAL HETZSCHOLDT

·

16 MAR

A Plea for Anti-Billionaire Regulation: Safeguarding Society, Civil Rights, and Democracy

Read full story