Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
Companies are rushing to set rules on how their employees can bet on prediction markets, as recent high-profile cases highlight how the platforms’ soaring popularity could open up lucrative opportunities for insider trading.
Helena Grannis, a partner specialising in corporate governance at law firm Cleary Gottlieb Steen & Hamilton, said the use of prediction markets such as Kalshi and Polymarket had rapidly become “a hot topic” for corporate boards.
She warned that standard insider trading policies designed for the stock market might not adequately cover the unique challenges posed by prediction markets, where staff may place wagers on issues that “don’t seem material — like how many times a chief executive uses a certain word in a speech — but [are] still confidential details”.
Prediction markets allow customers to bet on binary outcomes of future events and have grown rapidly by offering wagers on everything from sports results to election outcomes.
Bets have extended into the corporate realm, with punters now speculating on future stock prices, product launches, private company valuations and even the words used in corporate earnings calls.
Wagers include whether Oracle’s executive team will mention “AI” or “artificial intelligence” more than 50 times during the company’s earnings update, or which song will be the most played on Spotify in a given week.
The total value of all contracts bought and sold on these platforms climbed to more than $29bn in May, up from just $2bn in the same month last year, according to crypto data platform Artemis.
The increasing corporate focus on prediction markets comes on the heels of several high-profile legal cases. US prosecutors in April charged a soldier involved in planning the January raid to seize Venezuelan leader Nicolás Maduro with placing Polymarket wagers on the mission that netted more than $400,000. Gannon Ken Van Dyke, the soldier, has pleaded not guilty.
A Google software engineer, Michele Spagnuolo, allegedly used confidential information to make more than $1.2mn on Polymarket, according to criminal charges unsealed in New York federal court in May.
Spagnuolo, an Italian who lives in Switzerland, did not enter a plea and has been released on $2.25mn bail. Prosecutors allege he used an internal software tool to access confidential data and wager “substantial sums” on the top names users searched for in 2025. Google described the charges as “a serious breach of our policies.”
Meanwhile, the platforms themselves have tightened their own rules over the use of confidential information. Earlier this year, Kalshi fined and suspended an employee of the media company Beast Industries, founded by YouTube creator MrBeast, for placing bets using alleged inside information. Kalshi did not share the wagers placed. However, the platform enables users to bet on the content of MrBeast videos, such as the words MrBeast — whose real name is Jimmy Donaldson — will use in upcoming videos.
Kalshi said earlier this week that it would require customers seeking to bet on markets linked to material non-public information to disclose where they work in an online form.
Bans or restrictions on Kalshi and Polymarket by countries including the UK, Brazil and Spain could in any case supersede company policy. But corporate approaches to the fast-developing platforms have ranged from outright bans to restrictions around trading on company or client information. Other organisations are holding back from setting specific guidelines as they wait to see “how the regulatory landscape lands,” said Grannis.
US investment group Federated Hermes earlier this year implemented some of the strictest regulations for staff, with a blanket ban that prohibits its employees from even holding a prediction market account.
Explaining its policy, the company noted that prediction market trades based on individual stocks’ performance or events related to financial markets could “be viewed as a conflict of interest” with the transactions it undertakes on behalf of its funds or clients.
Federated Hermes also cited the “reputational risk” arising from concerns that staff could be perceived as trading on confidential information, and said that because most prediction market platforms do not produce account statements, employee transactions “cannot be pre-cleared and monitored like securities transactions in a brokerage account”.
The company does allow its employees to trade securities — including its own — provided they do not use non-public information or trade in designated closed periods.
Other companies have chosen to explicitly include the platforms in their anti-insider trading policies. As of March, cyber security company SentinelOne had updated its compliance rules to expressly include “placing bets or wagers on prediction markets based on material non-public information.”
Some, meanwhile, have fallen back on existing policies. Goldman Sachs said it already “strictly prohibit[s] trading on any material non-public information in all markets,” while cyber security company CrowdStrike, whose financial performance has been the subject of bets on Polymarket, said its “existing policies have always prohibited this type of behaviour”.
Amanda Litman, founder of non-profit organisation Run For Something, said she has banned staff from trading on any electoral race or candidate that the non-profit was “actively involved in supporting, monitoring, or considering for endorsement”.
Non-profit journalism organisation ProPublica in April announced that it had banned its staff from wagering on any “news event”, regardless of whether they are involved in coverage.
ProPublica assistant managing editor Diego Sorbara said the organisation stopped short of a blanket ban but still felt that setting some kind of policy was essential: “We basically felt it was a bad look.”
