Tarek Mansour says prediction market crime happens in plain sight

Tarek Mansour walked into DC Fintech Week with an argument that sounds paradoxical until you think about how markets actually clear: the more visible the transaction, the harder the fraud.

The Kalshi CEO's defence of prediction market transparency came at a moment when the sector has produced three notable enforcement cases in a single year, which his critics might read as evidence of a problem and which Mansour reads as evidence of a system working. Both of those interpretations are available from the same set of facts. Only one of them is correct.

Start with George Santos. The CFTC ordered him to return profits and pay a penalty for trading contracts on whether he would attend the State of the Union, while making public statements designed to move those contracts in his favour. Kalshi subsequently banned him permanently. What Mansour's team has disclosed is that roughly one hundred market participants flagged the suspicious activity within minutes of it occurring. That is not a compliance system catching a problem — that is a crowd catching a problem and handing it to a compliance system. The distinction matters, because it tells you something about where the actual oversight lives in a liquid prediction market.

The Gabriel Perez case is structurally different. A White House teleprompter operator with advance knowledge of speech content trading on word-frequency contracts is an information asymmetry problem of a kind that no market design fully solves. Polymarket faced something similar in the Van Dyke matter, where federal prosecutors alleged a Special Forces sergeant used classified intelligence about the Maduro capture operation to generate substantial profits on geopolitical contracts. The Perez disgorgement and the Van Dyke prosecution are not the same story — one is regulatory, one is criminal — but they share the same underlying shape: someone with state-adjacent information standing at the front of a queue the market doesn't know exists.

Mansour's response to this category of risk is mandatory employer disclosures and additional compliance checks for sensitive contracts. I find those measures more useful as deterrents than as detection tools. You cannot verify what someone's employer does not know they are trading. The honest version of his argument is that transparency makes post-hoc enforcement faster and cheaper, not that it makes the underlying information asymmetry disappear. He is right about the first part.

What I think is underpriced by the commentary around these cases is how much the enforcement record actually supports federal oversight of prediction markets rather than undermining it. Every action described — Santos, Perez, Van Dyke — involved the CFTC or federal prosecutors. The states arguing that sports prediction contracts belong under state gambling licensing have not produced a parallel enforcement record. New York has prohibited state officials from using nonpublic information for prediction market profit, which is a real policy step, but the regulatory infrastructure that actually caught and punished three separate bad actors this year is the federal one.