Trump’s speechwriter’s brother made $100,000 by trading prediction markets using insider information! With his aide gone, how did Kalshi find out?

For 10 years, you helped run the teleprompter for Trump at the White House. Not only did you know what the president was going to say, but you also knew what he would skip.
(Background: Kalshi lobbied for Trump through a group of former Trump aides and a lobbying crew for gamblers! Casino operators fired first, and Polymarket was targeted by Congress the same day.)
(Additional background: The U.S. House launches a thorough investigation into insider trading at prediction markets Polymarket and Kalshi, to prevent officials from profiting using “classified information.”)

Table of contents

Toggle

• How does a teleprompter staffer bet?
• How was it discovered?
• The “insider trading” problem in prediction markets
• Implications for Taiwan

White House staffer Gabriel Perez has been responsible for operating President Trump’s teleprompter since 2016. Recently, ABC News revealed that he used his position to bet on the Kalshi prediction market, earning more than $100,000 from speeches in over a dozen sessions.

How does a teleprompter staffer bet?

According to ABC News, what Gabriel Perez bet on wasn’t who would win—it was which specific words, phrases, or themes would be mentioned in Trump’s speeches. This is Kalshi’s “Mentions” market, where users can place bets on whether particular content will appear in a given public speech.

Perez’s advantage was straightforward: he knew the original version of the teleprompter script. When Trump skipped the prepared sections, Perez would close out his position mid-speech. ABC cited insiders saying this playbook was repeated more than ten times within about three months, covering addresses including intelligence briefings and the World Economic Forum.

How was it discovered?

Kalshi’s monitoring system detected abnormal trading patterns and proactively routed the trading data to the U.S. Commodity Futures Trading Commission (CFTC). Regulators then launched an investigation, tracing more betting footprints back to speeches.

According to reports, Perez is currently in settlement negotiations with federal regulators, and the White House has placed him on unpaid administrative leave. White House spokesperson Karoline Leavitt said Trump called the behavior “disgraceful.”

The “insider trading” problem in prediction markets

The growth trend of prediction markets is also amplifying the issue. According to CoinGecko data, this year’s first quarter prediction market trading volume hit a new high. Even if the overall crypto market is weak, growth momentum can’t be stopped.

Recent similar cases keep coming:

• March: Six Polymarket traders correctly predicted U.S. airstrikes on Iran before the end of February, earning about $1 million
• DeFi platform Axiom’s on-chain investigation: after blockchain sleuth ZachXBT published the allegations, several wallets had already placed bets, earning $1.2 million
• Venezuela: A trader correctly placed a bet before Maduro was captured, earning about $400,000

These developments have prompted lawmakers to pay closer attention. Last month, Representative Bryan Steil introduced legislation to ban members of Congress and their immediate family members from trading prediction market contracts tied to public policy and political outcomes.

Implications for Taiwan

Does Taiwan have prediction markets? In practice, Taiwan’s lottery and sports betting markets have existed for a long time, but “event contract”-type prediction markets (such as Kalshi and Polymarket) are still relatively new in Taiwan. If teleprompter staff can even profit from inside information, the information gap ordinary people may face when placing bets is even harder to imagine.

The competitiveness of prediction markets lies in “price discovery”—using market mechanisms instead of polls. But when inside information becomes part of the game, the reliability of price discovery declines. Gabriel Perez’s case is a perfect example that what you’re betting on matters more than where you’re betting.