Prediction markets have a known vulnerability: what happens when the person the contract is about also happens to be trading it? George Santos answered that question in the most George Santos way imaginable.

Santos, the former Republican congressman from New York whose tenure ended in expulsion over a cascade of fraud and fabrication allegations, reportedly traded against his own attendance at President Donald Trump’s State of the Union address. He had publicly stated he planned to attend. The market believed him. He then missed the event, blamed an airport delay, and walked away with more than $17,500 in profit.

How the trade worked

Kalshi, a prediction market platform regulated by the Commodity Futures Trading Commission, listed a contract on whether Santos would show up to the State of the Union. His public declarations that he intended to be there pushed the implied probability of attendance to roughly 75%.

In plain terms: the market was pricing in a three-in-four chance he would walk through the door. Santos, who apparently had better information about his own schedule than the market did, took the other side of that bet.

When he failed to appear, citing an airport delay due to weather, the contract resolved against attendance and Santos collected. His lawyer later argued there was no intent to deceive, framing it as a case of changing travel plans rather than deliberate manipulation.

Kalshi’s surveillance systems read it differently. The platform flagged the discrepancy between Santos’s public statements and his trading behavior, froze his account, and referred the matter to the CFTC and the Department of Justice in early June 2026.

The CFTC settlement

On July 31, 2026, the CFTC announced a settlement with Santos. He agreed to pay $35,000, a figure that includes forfeiture of his trading profits, and accepted a three-year ban from participating in prediction markets. He did not admit wrongdoing.

The case is a tidy illustration of what regulators call “informed trading” in its most literal form. Santos wasn’t working from leaked documents or insider tips. He was the information. He knew whether he planned to get on a plane, and the market didn’t.

Why this matters beyond the obvious punchline

Prediction markets derive their value from aggregating dispersed information into prices. That mechanism breaks down when the subject of a contract can trade on knowledge that is, by definition, unavailable to anyone else. It’s less like insider trading in the traditional sense and more like a boxer betting against himself. The information edge isn’t stolen. It’s inherent.

Kalshi’s decision to list a contract on Santos’s attendance created exactly this exposure. Whether he would show up was knowable only to Santos and perhaps his travel agent. Every other market participant was effectively guessing. The 75% implied probability reflects the market’s best read of his public statements, which Santos had every incentive to inflate if he intended to trade the other side.

The harder question is whether platforms should list contracts where the subject is also a likely participant in the market. Traditional financial markets have extensive disclosure regimes around this. A CEO trading their own company’s stock triggers specific reporting requirements under securities law. Prediction markets, still relatively young as a regulated asset class, are working out the equivalent framework in real time.

The CFTC’s willingness to pursue and settle this case, even for a relatively small dollar amount, signals that the agency considers self-referential trading a live enforcement priority. For platforms considering similar contracts, whether on politicians, athletes, or public figures who might plausibly trade them, the Santos settlement functions as a compliance benchmark. Surveillance systems need to flag when a contract’s subject is also an account holder.