Debate grows over whether US ownership stakes would control AI firms

In short: U.S. policymakers are debating government ownership in AI companies, and experts say the details decide whether it increases control or instead weakens oversight.

What's going on

Some proposals would give the federal government a small ownership stake in major AI companies but no vote and no seat on the board. A board is the group that helps run a company, like a steering committee. Reporting has described a Trump-style idea as voluntary and non-voting, and OpenAI has reportedly floated a 5% stake that also appears to be a minority stake without special control.

Other proposals are designed to give the government direct influence inside the companies. Senator Bernie Sanders has proposed a plan where major AI firms would pay a one-time tax in stock, and the public would hold voting shares plus equal seats on each company’s board. Voting shares matter because they can decide leadership and key decisions, like having a vote in a homeowners association.

Experts say it is not accurate to treat all government ownership as the same. A non-voting stake usually does not let the government directly tell a company what to do. But it can still change behavior, since regulators may hesitate to crack down on a company if the government also benefits when the company’s value rises.

What to watch

The key question is whether any proposal includes voting rights, board seats, or special veto powers. Those features would clearly expand direct government control. Without them, critics warn the bigger risk is conflicted oversight, where the referee also owns part of the team.

Source: NYTimes