LineMeta is building a standalone prediction market app, internally codenamed Arena — and according to reporting from NPR, Mark Zuckerberg first met with Kalshi's CEO about an outright acquisition before deciding to build in-house instead.
Let that sink in: the company with four billion users looked at prediction markets, tried to buy the market leader, got turned down, and decided the category was worth building from scratch anyway.
What Arena actually is
Arena won't take real money — at least not at first. Users will wager play-money points, video-game style, on news events and trending topics. Per internal documents reviewed by NPR, Meta plans to use its Llama AI to generate market questions automatically from trending topics, recommend personalized markets to each user, and resolve outcomes in near real-time.
The app will run separately from Facebook and Instagram, but it's designed to feed off Meta's enormous audience. And Meta has notably not ruled out real-money trading down the road. There's precedent for the interest: Meta already struck a deal in March to integrate Kalshi markets directly into Threads.
What it means for Kalshi and Polymarket traders
Short version: this is bullish for everyone in the ecosystem — including you.
More traders are coming. Arena's play-money model is a funnel. Millions of people will learn what a prediction market is by betting points on Meta, and a slice of them will graduate to real-money venues like Kalshi and Polymarket. The whole pool gets deeper.
More venues means more disagreement. Every new platform pricing the same events is another chance for prices to diverge. Kalshi vs Polymarket gaps are already a daily occurrence — add more venues and the map of mispricings only grows.
AI-resolved markets will be messy. Meta wants Llama to both create and resolve markets automatically. Anyone who's traded through a disputed resolution knows how hard that is to get right — expect controversy, and expect the sharps to profit from it.