Evening Update: Pentagon AI Chief Cashed In Millions on xAI Deal Amid Defense Contracts.

As defense contracts flow, millions follow—and ethics questions come knocking.

Small Bites is a reader-supported publication. To receive new posts and support my work, consider becoming a paid subscriber.

Hey Small Biters,

A senior defense official overseeing artificial intelligence policy holds stock in an AI company. The Pentagon signs deals with that company. The stock skyrockets. The official sells. Millions are made and then, almost as an afterthought, the ethics questions arrive.

The numbers alone tell a story that doesn’t need embellishment. What began as a stake valued between $500,000 and $1 million ballooned into a payout somewhere between $5 million and $25 million.

A gain not just impressive—but politically combustible. At the center of it is Emil Michael, the Pentagon’s under secretary for research and engineering. His role is not peripheral. It is central. He helps shape how the military adopts artificial intelligence.

He also held a financial interest in one of the companies benefiting from that push.

That company is xAI, founded by Elon Musk. Its flagship product, Grok, is part of the Pentagon’s growing AI ecosystem and during the period in which Michael held his stake, the Pentagon entered into not one—but two agreements involving the company.

Timing, in politics and ethics, is everything. The first agreement came in mid-2025, positioning xAI as one of several providers helping the military expand its use of artificial intelligence. The second arrived in December, just days after Michael received official notice that he would need to divest his holdings.

He did not sell immediately. Instead, the deal was announced first. Only weeks later, in January, did the sale occur.

To be clear, financial disclosure rules allow ranges rather than precise figures, leaving the exact profit somewhat obscured. But even within those ranges, the scale is unmistakable.

Hundreds of thousands turned into millions and in Washington, that transformation rarely goes unnoticed. Ethics experts have already begun weighing in. The concern is not subtle. Federal law prohibits government officials from participating in decisions that could benefit their own financial interests.

It is a bright line. Or at least, it is supposed to be. Richard Painter, a former White House ethics lawyer, did not mince words. Situations like this, he suggested, carry a high risk of violating criminal statutes.

The Pentagon, for its part, has defended the process. Officials insist that all rules were followed, that disclosure requirements were met, and that safeguards were in place.

The phrase “full compliance” has been deployed. But compliance, in cases like this, often becomes a matter of interpretation rather than clarity. Complicating matters further is Michael’s public posture. He has not operated quietly in the background. Instead, he has been an outspoken and sometimes combative figure in the Pentagon’s broader AI battles.

In disputes with other companies—most notably Anthropic—he has taken to social media to hurl personal insults at executives. Not exactly the demeanor of a neutral arbiter.

This matters because the Pentagon is not merely adopting AI. It is racing toward it. Posters line the halls urging personnel to embrace artificial intelligence, echoing wartime recruitment campaigns.

The message is clear: AI is not optional. It is the future and with that future comes contracts, influence, and money. Lots of money.

The scale of the Pentagon’s AI ambitions means that decisions made today will shape billions of dollars in spending tomorrow. Companies that secure early footholds stand to gain enormously.

Which makes any overlap between policy and profit more than just an optics issue. It becomes a structural concern. Michael’s background adds another layer. Before entering government, he built a career in the private sector, including a high-profile role at Uber. He is also known to have social ties to Musk.

None of this is illegal. But all of it contributes to a perception problem. Because in Washington, perception often matters as much as reality. The broader context only heightens the stakes. The Pentagon is already engaged in a contentious debate over how AI should be used—whether for surveillance, targeting, or autonomous weapons.

Some companies, like Anthropic, have drawn lines around those uses. Others have not and as those debates unfold, the individuals shaping policy carry immense influence.

Influence that, in this case, intersected directly with personal financial gain. The Pentagon insists its ethics framework is robust. Multi-layered. Protective. But frameworks are only as strong as their execution.

Ultimately, this story is not just about one official or one transaction. It is about a system grappling with a new frontier—where technology, power, and profit collide at unprecedented speed.

Artificial intelligence is rewriting warfare. It may also be rewriting the rules of accountability.

---

✍️

Gold in the wires,
profit in code,
policy written,
while fortunes explode.

The law may bend,
but trust may break,
and what is allowed,
is not always what we take.

---

🧭 A Small Bite to Carry

• A top Pentagon AI official made millions from xAI stock while overseeing decisions that affected the company.

• Ethics experts warn the timing and overlap could violate conflict-of-interest laws, despite official claims of compliance.

• As AI reshapes defense strategy, the line between public duty and private gain is becoming harder—and riskier—to ignore.

---