The short version: billionaire media founder Michael Bloomberg thinks the U.S. plan to let the government take a stake in AI companies would make “cronyism” more likely. In a Bloomberg Opinion op-ed published Monday, Bloomberg argued that the proposal would turn Washington from an industry regulator into an investor, giving officials financial incentives that do not belong in AI oversight.

Bloomberg frames the risk as political economy dressed up as national strategy. He wrote, “Somewhere, Karl Marx is smiling,” and said the “propaganda possibilities would make George Orwell blush.” The thrust is clear: he believes government shareholding would blur the line between supervising an industry and steering it toward outcomes that benefit insiders, not the public.

To understand why this argument is getting attention, look at what the source says is happening underneath the debate. The “initial deal” behind the American AI boom, as described in the piece, is that private investors help finance the breakthroughs, take on the risk, and private companies initially own the benefits. Later, the plan is for those benefits to flow into public markets, while government regulation comes “after the fact.” In this model, regulation is meant to correct or manage externalities once the technology exists and the companies are already in motion.

The same “deal” is contrasted with China’s approach. According to the source, China’s bargain is that companies still compete for investment and customers, while the government provides the compute. The implication is that the state accelerates capability by supplying the expensive inputs, but competition for money and customers still happens at the company level.

The source then claims that bargain is “showing signs of collapse” on the U.S. side. As AI costs soar, Chinese competitors are gaining ground, and Washington increasingly treats AI as a national-security asset. That combination is what puts Trump’s proposal into focus: the president is “considering taking a governmental stake into AI companies.” The key point is not just symbolism. A government stake changes who gets to influence direction, timing, and capital access, even if the program is described as supporting innovation.

And while the source says both the populist left and the right, and the AI companies themselves, have praised the proposal, Bloomberg is the one major voice in this account warning of the political side effects. He argues Americans do not need their government to own AI companies to share in the technology’s gains. His specific example is that once companies go public, people can buy shares, letting markets distribute ownership and upside without government holding power.

Bloomberg also makes a second, more “public value” argument. He says consumers and businesses already benefit from AI applications like fraud detection, medical research, and bookkeeping. If AI drives enough economic growth, he argues it could generate more tax revenue that funds public services. In that framing, the public does not need a federal shareholder to get the benefits. Instead, if companies are not contributing enough, he says Washington should “fix the tax code to serve the public,” not buy the companies.

The op-ed’s capstone is a prediction about governance drift. Bloomberg ultimately predicts that federal shareholders will likely lead to corruption because the market will transform into a “smoke-filled backroom.” That is a different risk category than technical safety or performance. It is about incentives, conflict of interest, and how influence works when government money and equity become part of the operating structure.

For executives, boards, and investors, the strategic stake is straightforward: a government equity stake can reshape how decisions get made, not only how they are reviewed. Even if a program begins with national-security justifications, it can create pressure to align with the agendas of political stakeholders who also own the upside. If capital allocation becomes partially political, competitive dynamics can change for the whole sector, including who wins deals, how partnerships form, and how companies think about information sharing and compliance.

In other words, the Bloomberg warning is about the incentive ecosystem surrounding AI. If Washington becomes both regulator and investor, the industry may end up navigating a new reality where oversight and ownership incentives move together. Bloomberg’s critique asks whether that is worth it, and the source provides a clear answer from one prominent opponent: he believes it is the fastest route to cronyism, not better innovation.