WASHINGTON – The tip came not from a whistleblower or a rival but from Kalshi’s own algorithms. During what the platform describes as routine surveillance of customer onboarding activity, its systems flagged an unusual pattern: trades that moved too precisely, too consistently, in step with what only a handful of people inside the White House could have known in advance.
The trader, it turned out, was Gabriel Perez, a veteran teleprompter operator who had spent years positioning words in front of the president’s eyes before they left his mouth. On Wednesday, the White House confirmed Perez had been placed on unpaid administrative leave and suspended from his duties, following federal referral of the case to the Commodity Futures Trading Commission.
White House Press Secretary Karoline Leavitt confirmed the action without elaborating on its scope. Leavitt said Perez was placed on unpaid administrative leave on presidential orders. The amount Perez allegedly earned through the scheme ran to nearly $100,000, a figure that multiple outlets citing sources familiar with the matter confirmed Wednesday, with some reporting the total exceeded that threshold.
Kalshi, the CFTC-licensed prediction market platform where Perez allegedly placed the bets, said it moved quickly once the pattern became clear. “Our surveillance team promptly flagged and referred these trades to the CFTC after an exchange investigation,” said Robert DeNault, Kalshi’s head of enforcement. “We have been assisting regulators.” According to the South China Morning Post, citing sources familiar with the matter, Perez is fully cooperating with federal authorities.
The mechanism of the scheme, as described by officials and sources, exploited a structural advantage that few people hold: advance access to the text of presidential remarks before they are delivered. Teleprompter operators receive the script of speeches, press conferences, and remarks in advance in order to load and scroll the text at the appropriate pace. That window, the gap between when the words are locked in and when they reach the public, created what regulators now believe was a tradeable edge.
Prediction markets, which allow bettors to wager on the probability of real-world events, have grown dramatically under the second Trump administration. Platforms like Kalshi and its rival Polymarket offer contracts on everything from Federal Reserve rate decisions to electoral outcomes to the content of specific speeches. Kalshi’s rise has attracted scrutiny before: bettors allegedly gamed Spotify charts to profit on music streaming prediction contracts, prompting the platform to demand both exchanges remove its logo. The White House teleprompter case is a different category of problem, one tied directly to government access rather than platform manipulation.
Kalshi operates under CFTC oversight, which grants it legal standing in the United States that unregulated competitors lack. That regulatory infrastructure also gave Kalshi the legal obligation to refer suspicious activity to federal authorities, the same framework that, in this case, caught one of the president’s own staff.
Whether the conduct constitutes a crime under existing law is the question the CFTC is now examining. Insider trading as a legal concept was developed for securities markets, where the SEC holds jurisdiction. Prediction markets under CFTC supervision occupy a different statutory terrain. The agency has brought enforcement actions against platforms for operating without authorization, but prosecution of individual traders for using non-public information to gain advantage in prediction markets is relatively untested ground. Legal scholars consulted by multiple outlets this week noted the case may require the agency to rely on anti-manipulation provisions rather than the more familiar insider trading statutes.
The case also raises questions about vetting and access within the executive office. White House staff who handle advance materials for presidential appearances undergo standard security reviews, but those processes are designed to assess trustworthiness relative to classified information, not to screen for financial conflicts that might be exploited through emerging market structures that did not exist in their current form until recently. No federal ethics framework currently requires White House employees to disclose prediction market accounts the way they must disclose securities holdings.
Trump’s reaction was characteristically brisk. The suspension came on presidential orders, and Leavitt’s public statement signalled the White House wanted the matter treated as resolved internally. Whether the CFTC pursues formal charges independently of the administration’s timeline remains open. Multiple sources told The New York Times and NBC News that the investigation was at an early stage, with no charges filed as of Wednesday evening.
What is not yet known is how long Perez operated the scheme, which specific speeches or events generated the winning trades, and whether he acted alone or with knowledge passed to others. Kalshi’s enforcement head declined to specify which contracts drew the surveillance team’s attention or what statistical threshold triggered the referral. Perez’s employment history at the White House, his hiring process, and the access controls governing advance speech materials have not been disclosed.
The episode illustrates an asymmetry that has emerged as prediction markets have matured: the platforms now carry enough real-money volume that access to advance knowledge of events, the kind scattered across a government employing hundreds of thousands of people, creates financial incentives that have not been anticipated in either personnel policy or financial regulation. Perez is the first known White House staff member to face scrutiny over such conduct. He is unlikely to be the last person inside any institution to notice the gap.
