Nila Kartika Wati
In 2021, Sam Altman, the CEO of OpenAI, authored a seminal essay titled "Moore’s Law for Everything," which outlined a utopian vision of the mid-2020s. In this document, Altman predicted a period of unprecedented abundance where the cost of fundamental human needs—housing, food, healthcare, and education—would plummet, theoretically halving every two years due to the efficiencies of artificial intelligence. To manage the resulting societal shifts, Altman proposed a radical redistribution mechanism: the "American Equity Fund." This fund was envisioned as a state-managed program financed by taxes on land and corporate equity, providing every adult citizen with a recurring cash dividend.
By mid-2026, however, the trajectory of the AI industry has diverged significantly from this early roadmap. While the technological capabilities of large language models have indeed scaled at a pace reminiscent of Moore’s Law, the promised redistribution of wealth has not materialized as a statutory right. Instead of a government-mandated equity fund, the industry has birthed massive corporate-controlled foundations. These entities engage in discretionary philanthropy rather than the systematic distribution of dividends. As the valuation of companies like OpenAI approaches the trillion-dollar mark, a new and complex debate has emerged regarding "nationalization"—a term that is increasingly used by both the radical left and the populist right, albeit with vastly different objectives.
The July 2026 Turning Point: OpenAI’s Equity Offer
In early July 2026, reports surfaced that OpenAI had entered into discussions with the United States government regarding a voluntary transfer of a five percent equity stake in the company. Based on OpenAI’s March 2026 funding round, which valued the firm at approximately $852 billion, this stake would be worth roughly $42.6 billion. This move was not prompted by a financial crisis or a request for a bailout, but rather as a proactive gesture during a period of intense regulatory scrutiny.
This proposal coincides with a broader shift in the industry’s stance on government involvement. Dario Amodei, CEO of Anthropic, recently published a policy essay encouraging the federal government to take a more active role in macroeconomic planning and tax policy to mitigate the labor market disruptions caused by AI. Unlike the traditional "hands-off" approach of Silicon Valley, these leaders are signaling a willingness to integrate the state into their capital structures, a move that critics suggest is less about public benefit and more about securing political "insurance."
A Strange Coalition: From Bernie Sanders to Steve Bannon
The proposal for public ownership of AI has created one of the most unusual political alignments in modern American history. Senator Bernie Sanders has introduced legislation that would dwarf OpenAI’s five percent offer, calling for a 50 percent public stake in the largest AI firms, to be taxed directly from their equity. Sanders’ rationale is rooted in the "commons": because AI systems are trained on the collective data of humanity and built upon decades of taxpayer-funded research, he argues the public is the rightful owner of the resulting wealth.
Unexpectedly, this sentiment has found echoes on the right. Former President Donald Trump has floated the idea of government equity stakes in critical technology firms, drawing a direct line to the populist appeal of Sanders’ plan. Steve Bannon has been even more vocal, accusing AI laboratories of attempting to "buy off" the government with minority stakes. Bannon has demanded that the government seize half of these companies, arguing that the industry has "scraped" the intellectual property of the American people to build its massive valuations. When a coalition includes democratic socialists, MAGA populists, and the CEOs of the companies themselves, the underlying motives and the definition of "nationalization" require deep interrogation.
The Financial Mechanics of the "AI Bailout" Model
The term "nationalization" is historically associated with the government taking control of failing industries to prevent systemic collapse, such as the 2008 banking crisis or the 2009 automotive industry rescue. In those instances, public capital was the last resort for survival. The current AI scenario is the inverse: the industry is offering equity while ostensibly at its peak.
However, a closer look at the balance sheets suggests a more precarious reality. The massive expansion of data centers and the acquisition of high-end semiconductors are being financed through significant debt and a complex web of "circular vendor-financing." In these arrangements, a dominant player like Nvidia might invest in an AI startup, which then uses those funds to purchase Nvidia’s chips. This creates a cycle that inflates valuations on paper without necessarily reflecting end-user demand.
Financial analysts have drawn comparisons to the "related-party" transactions that obscured the health of companies like Enron and Lucent in the past. While Nvidia has denied using special-purpose entities to hide debt, market analysts remain concerned that the industry is leaning heavily on vendor-financed demand. If AI growth slows, these circular arrangements could collapse. In this context, offering the government a five percent stake today serves as a hedge. By making the government a shareholder, these firms ensure that Washington has a vested financial interest in their continued success, potentially softening future regulatory or antitrust actions.
Shifting Public Sentiment and the Shareholder Strategy
The push for public equity also serves a psychological purpose for an industry facing a "trust deficit." A 2026 Pew Research Center survey of over 5,000 U.S. adults revealed that 40 percent of the population expects AI to have a net negative impact on society, while only 16 percent expect a positive outcome. Skepticism is particularly high among adults under 30, nearly half of whom believe AI will be detrimental to the social fabric.
By rebranding the public as "shareholders," AI firms aim to shift this perception. The logic is simple: individuals are less likely to advocate for the dismantling or heavy regulation of an industry if they believe they are direct beneficiaries of its profits. However, as critics point out, a minority equity stake without governance rights does not equate to public control; it merely creates a "quiet shareholder" in the form of the federal government, which may then be less inclined to exercise its oversight duties.
The Precedent of the "Equity State"
The current foray into government equity stakes is not without precedent. Under the Trump administration, the U.S. government engaged in what the Council on Foreign Relations described as the most significant push into private industry equity since World War II. This portfolio now includes approximately 30 companies across sectors such as rare earth minerals, semiconductor manufacturing, quantum computing, and defense.
Companies like Intel, US Steel, and MP Materials have seen their stock prices surge following the announcement of government investment. For instance, USA Rare Earth’s valuation jumped significantly upon the announcement of a $1.6 billion government deal. This has created a new pattern on Wall Street, where analysts now curate "buy lists" based on which companies they expect the government to invest in next.
Unlike the 2008 bailouts, where the government had a clear exit strategy, these current investments are in "healthy" companies with no specified timeline for divestment. This represents a fundamental shift in the American economic model—a transition toward a "state-capitalist" hybrid where the government acts as a venture capitalist, but often without the transparency or accountability required of public institutions.
The American AI Sovereign Wealth Fund Act
The most concrete legislative attempt to formalize this relationship is Senator Bernie Sanders’ "American AI Sovereign Wealth Fund Act." This bill serves as a direct counterpoint to the voluntary "gifts" proposed by OpenAI.
Key Provisions of the Sanders Bill:
- Equity Tax: A one-time 50 percent tax paid in stock for any AI firm with more than $200 million in annual AI-related gross receipts.
- The Commission for Democratic AI: Management of the fund by a seven-member independent commission, nominated by the President and confirmed by the Senate.
- Direct Dividends: A requirement that the fund pays out five percent of its value annually as a direct cash dividend to every U.S. taxpayer.
- Social Investment: Channelling the remaining proceeds toward national priorities, specifically healthcare and housing infrastructure.
Sanders argues that because these models were "trained on the sum total of human knowledge," the wealth they generate belongs to the people. While the bill is unlikely to pass in its current form—even Norway’s Sovereign Wealth Fund, the Global Pension Fund, caps its ownership of any single company at 10 percent—it establishes a benchmark for what "true" nationalization might look like.
Historical Comparisons: Nuclear Power vs. the Internet
The debate over AI ownership often ignores the historical precedents of transformative technologies. Nuclear energy, for example, was never left to the "rough hands of capitalists" in its infancy. It was developed as a state program, and civilian nuclear power was built under a strict regime of public ownership and licensing.
Similarly, the foundations of the internet (ARPANET) were entirely taxpayer-funded through the Department of Defense. However, while the internet’s infrastructure eventually moved into private hands, the AI sector has seen a "private-first" development cycle that relies on public data and public research but keeps the governance and revenue entirely behind corporate walls.
The current proposals to hand back a five percent stake are viewed by some as a "reverse-engineered" version of these historical models—where private firms, having achieved market dominance through public resources, offer a "crumb" of the cake to avoid being treated like public utilities.
The Distinction Between Ownership and Power
The core of the debate lies in the distinction between "equity" and "power." Nationalization, in its traditional sense, implies that the public has a say in directing a company’s priorities—such as ensuring worker protections, environmental standards, or ethical guardrails.
A five percent or even a 50 percent equity stake sitting in a government-held fund does not automatically grant the public this power. If the stake is "inert"—meaning the government does not exercise voting rights or demand board seats—it is simply a financial instrument. Without enforceable conditions on how these companies treat workers facing displacement or how they manage the massive energy demands of data centers, "nationalization" remains a hollow term.
Conclusion: A Narrow Window for Accountability
As the United States government moves closer to becoming a major shareholder in the AI industry, the need for structural accountability has never been more critical. If government equity stakes are accepted as "voluntary gifts" on corporate terms, it sets a dangerous precedent for all future negotiations between the state and concentrated corporate power.
Civil society and policymakers are currently in a narrow window of opportunity. Before these equity arrangements become normalized as standard practice, there must be a demand for transparency and governance. Any public stake in AI should come with:
- Governance Rights: The ability for the public to influence the ethical trajectory of AI development.
- Enforceable Terms: Clear consequences if companies fail to meet standards regarding labor displacement or data privacy.
- Public Accounting: A transparent, continuously updated ledger of how these public assets are being managed and who is benefiting from the dividends.
The "Moore’s Law for Everything" future may yet arrive, but whether it results in a broad-based "American Equity Fund" or a reinforced corporate monopoly depends entirely on whether the public demands power, rather than just a small piece of the profit.
