In January 2026, Goldman Sachs’s CEO called prediction markets “super interesting” and met personally with the founders of the two biggest platforms in the space. Prediction markets are websites where people bet on whether a specific event will happen — a company’s ranking, an election outcome, a war ending — much like betting on a sports game, except the “game” can be almost anything.
By May, the mood had shifted. Federal regulators charged a Google employee, trading under the handle “AlphaRaccoon,” with using inside knowledge of his own company’s unpublished rankings to place winning bets — walking away with roughly $1.2 million. It was the first case of its kind involving a private company’s internal information.
The pattern didn’t stop there. Back in March, the White House had already warned staff not to use non-public government information on these platforms, after unusual bets appeared just before a public announcement about a pause in strikes on Iran. By summer, a Major League Baseball–regulator partnership was set up to monitor betting tied to player information, and lawmakers raised alarms about the same risk spreading to soldiers and government insiders.
By July, Goldman Sachs did what six months earlier would have seemed unthinkable: it banned its own employees from betting on anything related to companies, elections, or financial markets on these platforms. Sports and entertainment bets are still allowed. No law required this — the bank drew the line before regulators did.
This is what happens whenever a new game opens where knowing something first lets you win: the old rule-breaking doesn’t disappear just because the venue is new. It just takes time for someone to get caught, and for everyone else to realize they could be next.
My take: What’s the difference between predicting and betting?
Not advice. Just how I see it.