Federal investigators believe Gabriel Perez, the man who has operated President Donald Trump's teleprompter since 2016, used his early access to speech texts to place winning bets on a website called Kalshi, according to ABC News reporting confirmed by The Verge on July 16, 2026.

Perez is accused of betting on what Trump would say during more than a dozen events, including the February State of the Union address, a Medal of Honor ceremony, and remarks at the World Economic Forum. He reportedly won more than $100,000 in total. Kalshi runs what it calls "mentions" markets, where users bet on whether a specific person will say particular words during a public event. Traders wagered more than $800,000 on Trump mentions markets alone, betting on whether the president would say terms such as "Hormuz," "rigged election," or "fake" during his speeches. (The Verge, NPR Illinois)

The case comes down to a simple information gap. A teleprompter operator, by the nature of the job, sees the text of a prepared speech before it is delivered. Kalshi's mentions markets pay out based on whether specific words are actually spoken at the event. Someone who already knows the speech text has information that the public does not, and that information directly affects the outcome of the bet. That is the advantage federal investigators reportedly believe Perez exploited.

Kalshi's head of enforcement, Robert DeNault, said the company's surveillance team flagged the trades and referred them to the Commodity Futures Trading Commission, a federal agency that regulates certain types of financial markets, following an internal exchange investigation. DeNault stated that Kalshi has charged the individual and has been assisting regulators. CFTC spokesperson Zach Fulton said the agency could not confirm or deny any investigation. (The Verge)

According to ABC News, the CFTC and Perez have discussed settling the case. The terms under discussion include Perez returning his earnings and agreeing not to place similar trades in the future. Federal prosecutors declined to open a criminal investigation. (The Verge, ABC News)

Kalshi recently implemented rules requiring users to disclose their employment before betting on markets flagged for heightened insider or manipulation risk. The timing of that rule change relative to the Perez matter is not specified in the available reporting, but the existence of the disclosure requirement points to an exchange building its defenses for a type of product that barely existed as a regulatory concern two years ago.

The broader context here is the collision between prediction markets and insider information. Think of it like a sports bettor who somehow knows the final score before the game is played. Traditional stock and futures markets spent decades building legal and technical frameworks to address this kind of problem. Kalshi's mentions markets are a new type of product: they pay out not on economic data or commodity prices but on the specific words spoken by a public figure at a scheduled event. That structure creates an easy opening for insider trading, because the number of people who see a speech before it is delivered is small, traceable, and in many cases publicly known. A teleprompter operator is one such person. Speechwriters, White House communications staff, and senior advisors would be in similar positions.

What gives this case particular weight is not the dollar amount. One hundred thousand dollars is modest by the standards of insider trading enforcement. It is the clarity of the information edge that matters. In stock markets, proving that a trader had information the public did not and traded on it often requires extensive investigative work to reconstruct who knew what, when, and how they shared it. Here, the advantage is built into the job: the teleprompter operator has the text before the speech is delivered. The bet resolves on whether specific words appear in that text. The connection is direct.

The decision by federal prosecutors to decline a criminal investigation, while the CFTC pursues a civil settlement, suggests regulators see this as a market-integrity matter rather than a criminal fraud case. That is consistent with how the CFTC has historically handled similar conduct in traditional futures markets: civil penalties, returning profits, and trading bans, not prosecution.

For Kalshi, the fact that its own surveillance team identified and referred the trades is notable. It suggests that the exchange has built detection systems capable of catching suspicious trading patterns in a product category that is still in its early stages. Whether those systems were in place before or after the suspected trades occurred is a question the available reporting does not answer.

Prediction markets have spent the past two years moving from the fringe toward mainstream regulatory acceptance, with Kalshi's CFTC-regulated status as a central part of that shift. The Perez case will not derail that trajectory. But it does illustrate, in unusually clear form, the kind of integrity challenge that comes with offering markets whose outcomes can be known in advance by a small, identifiable group of people. Every regulated market confronts insider trading. The question for prediction markets is whether the products they offer create advantages that are too easy to exploit and too difficult to police.