A White House teleprompter operator was fined for trading presidential mention markets

On 28 August 2026 the Commodity Futures Trading Commission issued release 9289-26, ordering Gabriel Perez to disgorge $107,539.02 in profits and pay a $65,000 civil penalty, with a three-year trading ban. Perez was employed as a teleprompter operator for the White House.

Between December 2025 and February 2026 he read presidential speeches before they were delivered and traded presidential mention market contracts on KalshiEX — event contracts the CFTC describes as "reflecting words or phrases the President may use during his speeches".

The Commission found that he "misappropriated material, nonpublic information obtained through his federal government employment in order to trade event contracts… for his personal benefit", and that in doing so he breached "his duty of trust and confidence".

Most event contracts settle on something no single person controls: an inflation print, an election, a central bank decision. A mention market settles on the words in a speech, and a speech exists in written form, in advance, in the hands of a small number of people.

That is a different risk from the one most traders price. The question is not whether the market is efficient but whether everyone in it is working from the same information — and here a portion of the counterparties knew the answer before the market opened.

The CFTC has said separately, in its February advisory on prediction markets, that exchanges carry "an independent duty pursuant to the core principles of the Act to maintain audit trails, conduct surveillance, and enforce rules against prohibited practices". This case is what that duty is for.

It does not mean mention markets are rigged, and the enforcement action is evidence the surveillance works rather than that it does not. It does mean the informational edge in a contract written on human speech can sit with people who are not traders at all.

The practical reading: treat a market whose outcome is knowable in advance by a handful of named people as a different instrument from one that settles on a public statistic, and size accordingly.