The federal government becoming a shareholder in the company it is meant to oversee is not a novel regulatory framework—it is a category error dressed up as industrial policy.

Reports emerged this week that the Trump administration is exploring taking an equity position in OpenAI, the San Francisco company behind ChatGPT that recently converted from a nonprofit structure to a for-profit entity valued north of $300 billion. The rationale, according to people familiar with the discussions, centers on ensuring American dominance in artificial intelligence and preventing foreign adversaries—read: China—from gaining technological supremacy. The mechanism remains unclear: direct Treasury investment, a sovereign wealth fund vehicle, or some hybrid arrangement involving the Commerce Department.

The conflict problem

Governments invest in strategic industries all the time. France owns stakes in Renault and Air France-KLM. Singapore's Temasek holds positions across technology and finance. But those investments exist within parliamentary systems with robust disclosure requirements and, crucially, separate regulatory bodies. The United States has no such architecture for AI.

The same administration that would hold OpenAI equity is simultaneously responsible for antitrust enforcement through the DOJ, safety standards through NIST, export controls through Commerce, and national security reviews through CFIUS. Every regulatory decision affecting OpenAI—from data privacy rules to compute restrictions to merger approvals—would carry the taint of financial interest. The government would be negotiating with itself.

The valuation question

There is also the matter of what, exactly, the government would be buying. OpenAI's valuation assumes continued exponential growth in enterprise adoption, successful navigation of copyright litigation, and the absence of any catastrophic model failures. These are not certainties. The company burned through billions developing GPT-5 and faces intensifying competition from Anthropic, Google, and a resurgent open-source ecosystem. A government stake at current valuations could easily become a government loss—politically radioactive in ways that private investors can absorb but taxpayers cannot.

The administration's apparent logic—that strategic importance justifies ownership—proves too much. By that standard, Washington should hold equity in Nvidia, TSMC's Arizona fabs, and every major cloud provider. The result would not be industrial policy but industrial entanglement on a scale that makes the 2008 bank bailouts look like arm's-length transactions.

Our take

This trial balloon deserves to be shot down before it gains altitude. If the administration wants to ensure American AI leadership, it has tools: R&D funding, immigration reform for technical talent, streamlined permitting for data centers, and yes, thoughtful regulation that creates predictable rules rather than arbitrary enforcement. What it should not do is become a shareholder in the very companies those rules are meant to govern. The appearance of corruption is sometimes worse than corruption itself, and a government with equity upside in OpenAI's success has no business deciding what OpenAI is allowed to do.