Meta Eyes a $130B Market With a New App Called Arena

Zuckerberg quietly ordered a team to clone Polymarket, and the timing is no accident.

Mark Zuckerberg has directed Meta to build a prediction markets app, internally named “Arena,” that would let users bet on real-world outcomes much like Polymarket and Kalshi. The reason is simple: prediction markets are exploding, with combined trades jumping from $50 billion in 2025 to more than $130 billion already this year.

Picture this. You open a new app, no Facebook logo in sight, and start wagering points on whether your team wins Sunday, how long a speech runs, or what happens next in the news cycle. No cash, at least not yet. Just you, the crowd, and a leaderboard. That is the bet Zuckerberg is quietly placing inside Meta right now.

What Happened

According to a New York Times exclusive, Zuckerberg recently dispatched a small team to build a smartphone app modeled on Polymarket and Kalshi. Two employees with knowledge of the plans described the project, speaking anonymously because it is confidential.

Here is the twist that makes it interesting. Users would not wager real money. The app would likely run on a video-game-style points system instead, though one person said Meta has not ruled out real money betting down the line. The app is called “Arena,” and it would run independently from Facebook, Instagram, WhatsApp, and Messenger.

Insiders called the effort experimental but a top priority. The growth playbook is classic Meta: leverage those massive social audiences and funnel them toward the new app.

The Backstory

This is not Meta’s first rodeo with predictions. Back in 2020, it launched Forecast, a crowdsourced prediction app that asked people to guess at how the early Covid-19 pandemic would unfold. It ran on a points system and was framed as a way to pool collective knowledge. Meta shut it down in 2022.

Then the world changed. Prediction markets erupted into a full cultural phenomenon, popping up during major sports events and even the Golden Globes telecast. The money tells the story: Kalshi and Polymarket pulled a combined $50 billion in trades in 2025, and the total has already blown past $130 billion this year.

Meta also has a long, bumpy history with standalone apps. Under a 2019 team called “New Product Experimentation,” employees tried launching social apps for podcasts, travel, music, and matchmaking. Few gained traction. The core problem was always the same: getting people to find and download a new app is brutally hard.

The Plan

So why try again now? Because Meta executives believe their flagship apps are running out of room. With Facebook and Instagram leaning ever harder into video, there are fewer corners left inside them to test fresh product ideas. More than 3.56 billion people already visit a Meta app every day, a number so large it has people asking whether the platforms have simply maxed out.

Arena is one of a handful of experiments in the pipeline. Another, called Meta Photos, would use artificial intelligence to generate new kinds of media. The common thread is Zuckerberg hunting for the next wave of online behavior and building apps to ride it. Meta declined to comment.

The Business Model Angle

Here is the part founders should tattoo on their whiteboard. Prediction markets are a beautiful business because the operator takes a fee on every single bet. High volume plus a cut of each transaction equals a potentially enormous, compounding revenue stream. That is why FanDuel, DraftKings, the crypto exchange Gemini, and even Trump Media have all stampeded into the space.

Meta’s edge is not the mechanic. Anyone can build a betting interface. Meta’s edge is distribution. The same machine that turns attention into ad dollars, detailed in the Facebook business model, can quietly point billions of eyeballs at a brand-new app on day one. Startups spend years and fortunes buying users. Meta can summon them.

The lesson: when you already own the audience, the smartest move is rarely inventing a new category. It is spotting a fast-growing behavior someone else proved out, then bolting it onto distribution nobody can match. Follow the users, then bring the crowd with you.

The Risk

Now the honest counterpoint, and it is a big one. Prediction markets have invited heavy legal scrutiny, precisely because you can bet on almost anything, which opens the door to insider trading. In April, federal prosecutors charged a member of U.S. Special Forces with using confidential information to bet on a top-secret operation to capture Venezuelan president Nicolás Maduro. Prosecutors say the soldier cleared more than $400,000.

The regulator meant to watch all this, the Commodity Futures Trading Commission, is running on its smallest staff in years just as its workload balloons. And the reputational knives are out. Senator Richard Blumenthal slammed Meta’s plans, accusing the company of “profiting from addiction” and pushing two bills, the Kids Online Safety Act and the Prediction Markets Security and Integrity Act.

There is also the small matter that Arena might never ship. Meta insiders cautioned the app is still in development and could be shelved entirely. Distribution muscle is real, but Meta’s graveyard of dead standalone apps is real too.

Quick Questions

The Business Model Analyst Take

The real story here is not gambling. It is distribution as a weapon. Meta keeps losing the race to invent new things, but it almost never loses the race to scale a proven thing, because it can flood any app with users overnight. For founders, that is the uncomfortable truth: a category you pioneered can be copied by a giant who simply owns more attention than you ever will. Your moat cannot be the idea. It has to be something Meta’s audience cannot hand a competitor for free. Watch whether Arena ships, and watch who it crushes if it does.

Based on reporting by Mike Isaac and David Yaffe-Bellany for The New York Times.