OpenAI proposes that AI companies give equity stake to the U.S. government

What happened?

OpenAI, through its CEO Sam Altman, has proposed that leading artificial intelligence companies grant the U.S. government a 5% equity stake in their businesses, according to the Financial Times as reported by Engadget. The idea would be to channel these stakes into a U.S. sovereign wealth fund, similar to those in other countries, so that the nation directly benefits from the sector's growth. This is not the first time Altman has suggested a close link between the state and AI: in 2023 he already proposed creating a global oversight agency similar to the IAEA, and in 2024 he advocated for the government to directly invest in AI infrastructure. However, the novelty here is the direct equity stake, which would turn the government into a co-owner of leading companies.

Why is this important?

This proposal represents a radical shift in the relationship between the private AI sector and the government. Traditionally, tech companies have resisted state intervention, but Altman argues that AI has the potential to generate immense economic value and also existential risks, so the government should have a 'seat at the table' both to share profits and to oversee the safe development of the technology. Historically, the United States has avoided state ownership of tech companies, unlike countries such as China or Singapore. Altman's proposal could change that paradigm, especially as AI investment has surged: according to PitchBook data, AI startups raised over $50 billion in the U.S. alone in 2024, and OpenAI is seeking a valuation of up to $300 billion in its next funding round. A 5% stake at that valuation would represent $15 billion for the sovereign wealth fund.

Moreover, the proposal comes at a time of intense debate over AI regulation. While the European Union advances its AI Act and China implements strict controls, the U.S. has opted for a looser approach based on voluntary guidelines. Equity stakes could be a tool for the government to have direct influence without heavy regulation, but it also raises questions about conflicts of interest.

What consequences will it have?

The consequences are multiple and complex. On one hand, it could align companies' incentives with national interests, reducing the temptation to externalize risks. For example, if the government is a shareholder, it could push for more investment in safety and avoiding algorithmic biases. However, it also poses serious conflicts of interest: the government would be both regulator and shareholder, which could lead to favoritism or reduced competition. Smaller companies or startups without the resources to cede capital could be marginalized, further consolidating the power of tech giants like OpenAI, Google, Microsoft, and Anthropic.

Furthermore, the proposal faces legal and political hurdles. It would require legislation or complex executive agreements, and could clash with the U.S. tradition of free markets. The reaction of other players like Google, Microsoft, or Anthropic will be key; if they do not join, the initiative could fail or create a patchwork of regulations. According to the Financial Times, Altman has already discussed the idea with some lawmakers, but there is no bipartisan consensus. On the other hand, the proposal could accelerate the creation of a tech sovereign wealth fund, something some economists have suggested for the U.S. to compete with funds like Norway's (over $1.7 trillion) or Singapore's GIC (over $700 billion).

In terms of market impact, the mere discussion of this idea has already generated volatility in the shares of unlisted AI companies and sparked debates in investment forums. If implemented, it could change how these companies are valued, as government ownership could reduce perceived risk but also limit upside for private investors.

What should readers know?

• It's not a done deal: This is an initial proposal from OpenAI, not official policy. It needs broad debate and consensus in Congress and among companies.
• The 5% is not trivial: Given the valuation of companies like OpenAI (estimated in the tens of billions), 5% would represent a massive injection of public capital. For example, if OpenAI is worth $300 billion, 5% is $15 billion, equivalent to NASA's annual budget (2024: ~$25 billion).
• Risk of regulatory capture: The government being a shareholder could bias its regulatory decisions in favor of the invested companies, to the detriment of competition. This would be similar to what happens in countries with state-owned enterprises, like China, where the government favors its national champions.
• International precedents: Countries like Singapore or Norway have sovereign wealth funds that invest in technology, but not with such direct stakes in specific companies. Norway's fund, for example, invests in over 9,000 companies globally, but without controlling stakes. Altman's proposal is more akin to the 'golden shares' that some European governments hold in strategic companies.
• National security implications: AI is a strategic sector; giving the government a stake could be seen as a way to maintain control over dual-use technologies, preventing them from falling into adversaries' hands. This is relevant after the Biden administration's export restrictions on AI chips to China.
• Possible alternatives: Some experts suggest that instead of equity stakes, the government could establish a tax on AI profits or require mandatory licensing. Altman's proposal is just one of several options under discussion.

“If the government becomes a shareholder in the leading AI companies, it will have a direct financial incentive for those companies to succeed, which could conflict with its role as an impartial regulator.” — Analyst at TheVortiq

In conclusion, Altman's proposal is a bold move that could redefine AI governance in the U.S. and globally. However, its feasibility is uncertain and will depend on political will, industry reaction, and public opinion. The coming months will be crucial to see if this idea takes root or remains a mere trial balloon.