Mention markets face manipulation review as CFTC examines speech bets

A teleprompter operator who worked for President Trump placed bets on words the president would or would not say during public appearances. Kalshi's surveillance systems flagged the pattern and referred the matter to federal authorities. That single case has now pulled an entire contract category into formal regulatory review.

The CFTC is examining mention markets — contracts that resolve on whether a named public figure speaks a specific word during a broadcast or event. Kalshi has removed all sports-related versions from its platform while the inquiry runs. The contracts let users bet on whether a sportscaster would say "MVP" or "ankle" during a game. That sounds narrow. The structural problem is not narrow at all.

The manipulation standard that governs self-certified contracts in the US requires operators to confirm a given market is not "readily susceptible to manipulation." Mention markets tied to speech face a design-level tension with that standard. The person with information about what a speaker will or will not say does not need to be a sophisticated trader to exploit it. A teleprompter operator is not a hedge fund. The information asymmetry is almost perfectly one-sided, and the market mechanism has no natural way to price that asymmetry out.

Luana Lopes Lara, Kalshi's co-founder, has been an internal advocate for the category, seeing speech-based contracts on award shows, earnings calls, and reality television as a user acquisition channel beyond Kalshi's core sports audience. That is a coherent commercial argument. The problem is that earnings-call mention markets create the same asymmetry in a different register — anyone with advance knowledge of a CFO's prepared remarks holds a structural edge that no amount of liquidity can neutralise.

Sports wagers account for more than 80% of Kalshi's weekly trading volume, and mention markets tied to athletic events were a meaningful slice of that. The revenue consequence of removing them is real. The strategic consequence may be larger: Kalshi's self-certification framework depends on regulators accepting that the company's internal surveillance is sufficient to catch bad actors before harm compounds. The teleprompter case is precisely the kind of outcome that tests that premise — and the fact that Kalshi detected it and referred it is both a point in the company's favor and evidence that the detection came after the trades were placed.

Polymarket runs mention markets on its international platform, outside CFTC reach, and does not offer them domestically. That jurisdictional split looks like a deliberate hedge, and it is probably the correct one given where the regulatory pressure is landing.

The reporting attributes bipartisan skepticism to the CFTC's internal discussions. I think that framing understates the structural issue. The problem with mention markets is not political — it is mathematical. A contract whose resolution condition is entirely within the knowledge set of a specific, identifiable person is not a prediction market in any meaningful sense. It is a wagering instrument that rewards proximity to the source. The CFTC's manipulation-susceptibility standard exists precisely to exclude that category of contract, and the current review is the agency working out whether the existing guidance covers it or whether new guidance is required.

That determination will matter beyond mention markets. The same logic applies to any contract where one party holds non-public information that is directly decisive for resolution — drug trial insiders betting on Kalshi's cancer trial contracts, political staffers betting on legislation timing, earnings traders with material non-public information betting on corporate event contracts. Mention markets are the visible edge of a broader question about where the manipulation-susceptibility line sits, and the CFTC's answer will set the standard for the entire self-certification framework.