AI Models Now A New Playground For Online Gamblers

Gamblers are wagering millions on AI models like GPT-5 and Gemini, using social media, GitHub, and rankings to predict top performers.

Fans are placing big bets on AI models as if they were sports stars, as noted in a report by The Wall Street Journal (WSJ).

Prediction platforms Kalshi and Polymarket receive millions of dollars in monthly bets from gamblers who predict which AI model will perform the best. The AI arms race manifests across social media, ranking sites, and niche corners of the internet, creating opportunities for those looking to profit.

Foster McCoy, a 27-year-old day trader, made $10,000 in a few hours by betting against OpenAI’s GPT-5. “You’re just betting against what the other guy knows,” he says, crediting his success to “being chronically online,” as reported by WSJ.

McCoy operates his trading business from home using X, Discord and LMArena which provides a leaderboard rating system for AI models in blind tests. WSJ reports that since early 2025, he traded $3.2 million on Kalshi, earning $170,000.

Similarly, Harvard student Rishab Jain uses social media posts and GitHub repositories and source files to predict upcoming model releases. “I’m almost obsessively up-to-date with what’s going on in this world,” he says, as reported by WSJ. His efforts have earned him $3,500 since June.

Betting strategies vary: some focus on big AI players, others on lesser-known models or arbitrage opportunities between platforms. After GPT-5’s release, Elon Musk’s posts praising xAI Grok caused its “Grok to Win” market to spike over 500% before falling again, as noted by WSJ.

Trading volume on AI prediction markets has surged to roughly $20 million this month, ten times higher than at the start of 2025, a Kalshi spokesman says to WSJ. Each bet, or “contract,” reflects the odds, and bettors can cash out when the price rises.

Economics professor Robin Hanson notes to WSJ, “When you have better information in these kinds of markets, you can make better decisions. If you know a little more, you make more money.”

For now, AI betting attracts both sharks like McCoy and Jain, and casual players like James Cole, who says to WSJ, “I’m just a dude that likes tech and has some time and money. I’m speculating with mostly instinct and 10 minutes of research.”