Meta May Be Building Its Own Prediction Market App. Here’s What We Know.

Prediction markets have become some of the fastest-growing sites on the internet, enabling people to bet on real-world events and win money when they’re right. Platforms like Kalshi and Polymarket surged in popularity during the 2024 U.S. presidential election, and are now pulling in millions (or even billions) of dollars in trading volume a day. 

Like a lot of tech leaders, Mark Zuckerberg has taken notice — and he wants in. According to a New York Times report, the Meta CEO has directed a team to build a prediction market app that will operate independently from Facebook or Instagram. The platform has not been officially announced, but it is said to be a high-priority project at the company. Zuckerberg has also reportedly instructed executives to speak with Polymarket and Kalshi directly as Meta explores ways to replicate the format for its own users.

Known internally as “Arena,” the app would allow users to speculate on future events like any other prediction market platform. However, it is expected to initially launch with virtual currency instead of real-money wagering. Reports also suggest that Meta is looking to turn Arena into a more comprehensive social experience rather than simply another form of online betting.

Sources familiar with the project told the New York Times that Arena is being tested internally, but that it may never be released publicly. Even so, the details that have emerged thus far offer a glimpse into how Meta hopes to adapt the booming prediction market model for social media.

So, how exactly will Arena work, what are Meta’s plans for it and why is the company steering clear of real-money wagering — at least for now? Here’s what we know so far.

First, What Are Prediction Markets?

Prediction markets are platforms where people can buy and sell contracts tied to the outcome of specific future events, whether that be the winner of a sporting event, the box office performance of an upcoming movie, the outcome of an election or tomorrow’s weather. 

Users choose to buy either “Yes” or “No” shares, representing a binary outcome of an event occurring or not. The price of each share directly correlates to the market’s collective estimate of that outcome’s likelihood, ranging from $0.01 to $1. For example, if users believe there’s a 90 percent chance the Federal Reserve will raise interest rates at its next meeting, “Yes” contracts would trade for $0.90 and “No” contracts would trade for $0.10. Traders can sell their contracts before the market closes or hold them until the outcome is determined. Once the event is resolved, winning shares are worth $1 and losing shares are worth nothing. 

Similar to other forms of betting, prices in prediction markets constantly rise and fall depending on new information and changing expectations among traders. But unlike traditional sportsbooks or casinos, where people are betting against the “house,” prediction markets don't set odds or take the opposite side of users’ positions. Rather, participants trade directly with one another, allowing the market itself to set the value.

Related ReadingPrediction Markets Explained: How Event Contracts and Platforms Work

What Do We Know About Arena So Far?

Although Meta has not announced — or even publicly confirmed the existence of — Arena, a vague picture of how the app could work is beginning to take shape.

One of Arena’s biggest differentiators at the moment is that it apparently won’t use real money. According to the internal documents obtained by NPR, users will receive a “daily virtual allotment” of “play money” that can be used to bet on the “outcome of future events.” It’s still unclear whether Meta plans to introduce real-money trading later on, but reports indicate the company has discussed the possibility of doing so.

Internal documents also state that artificial intelligence will play a central role in how Arena operates. Meta’s own Llama language model will automatically generate questions from trending topics and make “personalized market recommendations” to individual users. Plus, the AI will resolve markets in “near real-time,” according to the documents. This would be a departure from established prediction markets like Polymarket and Kalshi, which rely on a combination of predefined rules, external data sources and, in some cases, human review or dispute resolution to settle contracts. If implemented as described, Arena’s AI would effectively act as the final arbiter of whether or not a given event happened, thus determining who wins and who doesn’t.

Meta employees told the New York Times that the target user base for Arena will be between the ages of 18 and 34, and that the company is aiming to reach at least 100 million monthly active users, or “predictors.” The app is expected to cover a wide range of topics (sports, politics, finance, entertainment, pop culture, etc.) and will reportedly be designed to facilitate social interaction as well as play betting, with the ultimate goal of encouraging users to share their opinions, debate current events and compete with each other.

“We believe that prediction markets are one of the more interesting new content types,” Ime Archibong, a VP of product at Meta who is leading the Arena initiative, said in an internal post later shared with the Times. “With the right containers, the social conversation is the payoff as people aim to show off how good they are at predicting things to their friends.”

Eventually, sources say Meta plans to integrate parts of Arena directly into Facebook, Instagram, Messenger and WhatsApp, allowing users to make wagers in group chats, share them in Stories and Reels and compare their performance through features like leaderboards. So, instead of existing as a standalone platform, Arena could become another way for people to engage with one another across Meta’s entire ecosystem — a concept Archibong called “betting as conversation.”

“When markets are built into a personalized feed, they stop feeling like something for traders,” he continued. “They start to feel like part of the conversation that is tied to memes, moments and whatever people are paying attention to. Betting becomes less about analysis and more about saying ‘This is what I think.’”

Another Platform to KnowWhat Is Polymarket? How the Prediction Market Platform Works

Inside The Rise of Prediction Markets

Meta is the latest in a long line of companies looking to capitalize on the white-hot prediction market industry, which is projected to be worth $1 trillion by the end of the decade. Major sports betting sites like DraftKings and FanDuel have launched prediction market services. Same goes for crypto exchanges like Coinbase, Kraken and Crypto.com. And just about every major trading platform is now offering some kind of prediction market-style product. 

At the same time though, prediction markets have been under intensifying legal and regulatory scrutiny in the United States, with critics arguing that contracts tied to elections, geopolitics, wars and other sensitive events more closely resemble market manipulation and insider trading  than traditional gambling. In one famous example, a U.S. Special Forces member was charged with misusing classified information about the then-secret mission to capture Venezuelan President Nicolás Maduro to win more than $400,000 on Polymarket.

The ethics of the industry have been criticized as well. Senator Richard Blumenthal, a Democrat representing Connecticut, called out Meta’s Arena plans in particular, claiming the business model is “profiting from addiction,” and encouraged people to support his legislative efforts to strengthen online child safety and prediction market rules. Similar concerns have apparently been expressed at Meta, too. According to the New York Times, internal message boards and private chats show some employees questioning whether Arena crosses an ethical line for one of the world’s largest social media companies, fearing that it could normalize risk-taking behavior or serve as a gateway to other forms of gambling.

Nevertheless, prediction markets have found a powerful ally in President Donald Trump. He has taken a more favorable stance toward them than past administrations, attacking states’ efforts to rein them in and calling for the Commodity Futures Trading Commission to retain “exclusive authority” over their regulation. The CFTC has since proposed a new set of rules for prediction markets, taking a generally more permissive approach while also giving the agency the authority to review and block contracts it deems aren’t in the public interest or that seem susceptible to manipulation.

For what it’s worth, the president and his family also have financial ties to the industry. In 2025, Truth Social, the social media app owned by Trump Media & Technology Group, released a prediction market product. Meanwhile, Donald Trump Jr. is reportedly an advisor to Kalshi and an investor in Polymarket.

More on MetaMeta Is Spending Billions to Bend America’s AI Policy to Its Will

Why Is Meta Getting Into Prediction Markets?

This is not the first time Meta is delving into the world of prediction markets. In the early days of the Covid-19 pandemic in 2020, the company released an app called Forecast, where people could use points to make guesses about what might happen next in the world, as well as share information with others. Meta shuttered Forecast two years later, but is now developing Arena as a “rebuild,” according to internal documents shared with NPR. 

Meta’s renewed interest in prediction markets is hardly surprising given how lucrative this business has become. Kalshi and Polymarket alone processed a combined $50 billion in trades in 2025 — and that trading volume has already more than tripled this year so far. In fact, Zuckerberg was reportedly in talks with Kalshi CEO Tarek Mansour to buy the startup last year, but the discussion ultimately fell apart.

At first glance, all of this explosive growth might make Meta’s decision to create a prediction market without any actual monetary stakes seem counterintuitive. But starting out with a play-money model could buy the company some time to gauge user interest without immediately having to deal with all the regulatory challenges of real-money trading. After all, there are dozens of pending lawsuits challenging the legality of prediction markets right now, and the CFTC is still trying to figure out how to effectively oversee the industry. It could take years for the dust to fully settle.

In the meantime, the primary purpose of Arena (should Meta choose to release it) is to drive engagement. More than 3.5 billion people use at least one of Meta apps every day, but the company has struggled to launch any successful standalone products in recent years. According to The New York Times, Arena is part of a broader push by Zuckerberg specifically to build new apps around emerging online behaviors, like events betting. This way, the platform serves as more than a fun game, but as another means of keeping users plugged into the Meta ecosystem.

Want to Learn More?Everything You Need to Know About Kalshi

Frequently Asked Questions