White House Aide Made $100K Betting on Trump Speeches

The Man Who Reads Trump’s Speeches First Just Became a Federal Case

Gabriel Perez has operated Donald Trump’s teleprompter since the 2016 campaign. He’s one of a small handful of aides trusted with the president’s prepared remarks before anyone else sees them, and he’s known to take last-minute edits directly from Trump himself, often minutes before a speech begins. That access is now at the center of a federal investigation: the Commodity Futures Trading Commission (CFTC) alleges Perez used his advance knowledge of speech content to place winning bets on Kalshi, the prediction-market platform, profiting close to $100,000 across more than a dozen speeches over roughly three months.

It’s the first known case of a White House employee accused of trading on inside information through a prediction market rather than a stock exchange, and it lands at a moment when regulators are already struggling to decide whether prediction markets are subject to insider-trading rules at all. Kalshi’s own surveillance systems caught the pattern before the government did, freezing roughly $90,000 of Perez’s winnings. The White House has placed him on unpaid leave, and Trump has called the situation a “disgrace.” What follows is how the alleged scheme worked, why it’s legally murkier than ordinary insider trading, and where this case fits into a much bigger fight over what prediction markets are allowed to be.

What Are “Mentions” Markets?

Kalshi is a federally regulated prediction market — technically a designated contract market under CFTC oversight — where users trade “event contracts” that pay out based on whether a specific real-world event happens. A subset of these, known as “Mentions” markets, let users bet on whether a public figure will say a particular word or phrase during a scheduled speech or appearance. Unlike a traditional sportsbook wager on a game outcome, a Mentions contract resolves based on the literal transcript of a specific, scheduled event — which makes it unusually sensitive to anyone with advance knowledge of what’s actually going to be said.

That sensitivity sits inside a broader regulatory gray zone. Since 2021, platforms like Kalshi have listed event contracts on categories far removed from traditional commodities or financial indices — video game release dates, awards-show winners, election outcomes, and now speech content. Sports contracts alone account for more than 85% of Kalshi’s trading volume, and federal courts are currently split on whether these products even count as regulated “swaps” under the Commodity Exchange Act, with two district courts ruling yes and two ruling no. The CFTC’s core anti-fraud authority over conduct like Perez’s is Rule 180.1 — a rule modeled on the SEC’s classic insider-trading provision, Rule 10b-5, but written for commodities and event contracts rather than corporate securities. Whether that securities-derived framework cleanly applies to someone betting on words in a presidential speech is, as of this case, still an open legal question.

Who Is Gabriel Perez, and What Access Did He Have?

Perez’s role is officially described as technical assistant to the president, but functionally he is the last set of eyes on nearly all of Trump’s prepared remarks before delivery. Reporting indicates he routinely receives last-minute edits directly from Trump — meaning in some cases he had more current knowledge of a speech’s exact wording than even the staff who drafted it. That combination — operational access to the literal text of an upcoming, market-relevant event, held by someone with no formal restriction on personal prediction-market trading — is precisely the vulnerability Mentions markets are built to exploit, whether the trader intends to exploit it or not.

How the Alleged Scheme Worked

The Trades

CFTC investigators identified bets across more than a dozen Trump speeches over a roughly three-month window, including a December primetime address, Trump’s January remarks at the World Economic Forum in Davos, Switzerland, and a March Medal of Honor ceremony. Reported profit figures vary slightly by outlet — CFTC-linked reporting cites more than $90,000 in frozen winnings, while some outlets describe the total as “nearly $100,000” — but the range is consistent across sources: a sustained, repeated pattern of wins on a market where the specific wording of a speech is the entire basis for the payout.

How Kalshi Caught It

Perez wasn’t caught by government auditors reviewing his personal finances — he was caught by the platform’s own trade-surveillance systems, which flagged betting patterns on presidential “mention” markets that didn’t resemble typical trading behavior. When Kalshi’s investigators traced the account, they found it belonged to a federal employee. That detection method matters: it means the case surfaced through private-sector market-integrity monitoring rather than proactive government ethics screening of White House staff’s personal trading activity, which raises its own question about how many similar cases might go undetected without a platform choosing to look.

Not Kalshi’s First Insider-Trading Case

This isn’t the first time Kalshi has flagged and penalized this kind of behavior. A CFTC advisory issued in February 2026 referenced two earlier Kalshi enforcement actions: one involving a political candidate who traded event contracts tied to his own candidacy, and another involving an employee of a company affiliated with a YouTube channel who traded contracts tied to that channel’s own video releases. In both cases, Kalshi imposed financial penalties and suspended the traders — a private, platform-level enforcement pattern the Perez case now follows, just with a considerably higher public profile.

The Legal Gray Zone

Classic insider-trading law exists to protect investors trading a company’s securities from being disadvantaged by someone with material, nonpublic corporate information — a framework built around shareholders, corporations, and tradeable financial instruments. A Kalshi Mentions contract on a presidential speech doesn’t fit that mold cleanly: there’s no company, no shareholders, and arguably no traditional “victim” class being defrauded in the way a securities-fraud case requires. The CFTC’s February 2026 advisory itself acknowledged this uncertainty, suggesting that trading event contracts on material nonpublic information might violate Rule 180.1 even without the kind of “misappropriation” relationship that traditional insider-trading cases require — but conceding that how far that theory extends is still legally unsettled.

That ambiguity is exactly why the CFTC is reportedly pursuing a negotiated settlement with Perez rather than pressing a novel legal theory through contested litigation. Settlement terms reportedly under discussion would require Perez to surrender his profits and agree not to make similar trades again — a resolution that lets regulators claim enforcement without forcing a court to rule on whether Rule 180.1 actually reaches this fact pattern.

Is This Really “Insider Trading,” or Just an Occupational Information Edge?

There’s a genuine argument that this case doesn’t resemble insider trading so much as a more mundane kind of unfair advantage — closer to a blackjack player counting cards than a corporate executive dumping shares before bad news breaks. Perez wasn’t extracting confidential information from a company he had a fiduciary duty toward; he was using the ordinary, sanctioned knowledge that came with doing his job, on a betting product that happens to key off exactly the information his job required him to know. On this reading, the failure isn’t really Perez’s judgment — it’s that nobody built a compliance wall between “White House staff with advance knowledge of presidential remarks” and “publicly tradeable markets that pay out based on those remarks” before Kalshi’s Mentions product existed.

The counter to that argument is that motive and mechanism don’t change the effect: whether Perez thought of himself as exploiting an information asymmetry or simply making informed bets, he was systematically betting against counterparties who structurally could not know what he knew, on a market specifically designed to price that exact uncertainty. Kalshi’s own prior enforcement actions — against a candidate betting on his own election and an employee betting on his own company’s content — suggest the platform itself doesn’t accept the “it’s just occupational knowledge” defense, regardless of whether Rule 180.1’s securities-law ancestry maps perfectly onto this new asset class. The disagreement here isn’t really about the facts; it’s about whether prediction markets should import securities law’s insider-trading framework wholesale, adapt it, or build something new specifically for event contracts — and that’s a question regulators, courts and Congress haven’t answered yet.

Data & Evidence Summary

Methodology note: figures and details are compiled from CFTC-sourced reporting relayed by ABC News, CBS News, CNN, NPR and CNBC. Profit figures differ slightly ($90,000 frozen vs. “nearly $100,000” reported total) likely because frozen funds and total realized winnings across the full trading period aren’t necessarily the same number; this piece has not independently reconciled the two.

Implications

For prediction-market platforms, this case is likely to accelerate compliance requirements around “insider” trading categories — expect Kalshi and competitors like Polymarket to build more explicit restrictions (or disclosure requirements) around event contracts tied to figures whose staff, family, or associates could plausibly have material advance knowledge of the outcome.

For federal ethics policy, the case exposes a genuine gap: government employees are subject to extensive financial-disclosure and conflict-of-interest rules around traditional securities, but prediction-market betting on event outcomes tied to their own official duties appears to have fallen outside existing ethics training and screening, at least in this instance.

For the broader prediction-market industry, this adds another high-profile data point to the argument, already being made by Democrats in Congress since at least April 2026, that Kalshi-style platforms need clearer federal rules — not just platform-level self-policing — around both sports betting overlap and insider-information misuse, a debate already playing out over Kalshi’s disputed jurisdictional status in multiple states.

Counterpoints and Limitations

No criminal charges have been reported in this matter — this is, at least publicly, a CFTC civil enforcement matter proceeding toward a negotiated settlement, not a Department of Justice criminal prosecution. Readers should not assume this case will result in criminal liability; a civil settlement with disgorgement of profits is a meaningfully different outcome than a criminal insider-trading conviction.

The exact scope of Rule 180.1’s applicability to event contracts remains legally unsettled, and because this case appears headed toward settlement rather than litigation, it’s unlikely to produce a binding court ruling clarifying that question. Future similar cases may still face genuine legal uncertainty about whether this conduct is enforceable at all.

Finally, the precise profit figure is reported inconsistently across outlets (roughly $90,000 frozen versus “nearly $100,000” in total alleged gains), and this article has not been able to verify which figure reflects total realized profit versus the specific amount Kalshi froze.

Conclusion

Whether or not Gabriel Perez set out to exploit his access to Trump’s speeches, the case exposes a compliance blind spot that prediction markets have created almost by accident: a betting product whose entire value depends on information that a small number of government employees routinely possess as a normal part of their jobs. Kalshi caught this one through its own surveillance, and the CFTC appears headed toward a settlement rather than a legal showdown over whether century-old insider-trading principles actually reach a market this new. That unresolved question — not the size of Perez’s winnings — is likely to be the more consequential legacy of this case, shaping how prediction markets, government ethics offices, and Congress handle the next version of this problem before it happens again.

FAQ

Who is Gabriel Perez and what is he accused of?
Perez is a White House technical assistant who has operated Donald Trump’s teleprompter since 2016 and typically has final access to his prepared remarks. He’s accused of using that advance knowledge to bet on Kalshi’s “Mentions” markets, profiting close to $100,000 across more than a dozen speeches.

What are Kalshi “Mentions” markets?
They’re event contracts that pay out based on whether a public figure says a specific word or phrase during a scheduled speech or appearance, resolved against the literal transcript of the event.

Has Gabriel Perez been criminally charged?
No criminal charges have been publicly reported. The CFTC is reportedly pursuing a civil settlement that would require Perez to return his profits and stop making similar trades.

Is betting on prediction markets using inside information actually illegal?
It’s legally uncertain. The CFTC’s Rule 180.1, the main tool being applied here, was modeled on securities insider-trading law but wasn’t originally built for event contracts on things like speeches, and courts have not yet ruled definitively on how far it extends to this kind of case.

How did Kalshi discover the alleged scheme?
Kalshi’s internal trade-surveillance systems flagged unusual betting patterns on presidential “mention” markets; when the company investigated the accounts involved, it found the trader was a federal employee.