Meta Builds a Prediction Market App Called Arena as Zuckerberg Chases the Category

Mark Zuckerberg has told Meta staff to build a standalone prediction market app, code-named Arena, that would let users wager virtual points on the outcome of sports games, elections, and political decisions. The New York Times first reported the plan on June 23, and Meta has confirmed the project is in development. Arena would sit apart from Facebook, Instagram, WhatsApp, and Messenger as its own product, though those apps could funnel users into it. The reporting arrives as Kalshi closes a funding round at a roughly $40 billion valuation, the clearest sign yet that event contracts have moved from a fringe product into a category large incumbents now want to own.

Arena starts on play money, not cash. Users would receive a daily virtual allotment of points to bet on future events and earn more by guessing correctly, a game-like loop closer to a mobile title than to a brokerage. Meta has not ruled out adding real-money trading later, according to the reporting, but the initial build avoids the regulatory machinery that real event contracts require in the United States. Whether Arena would eventually plug into a Commodity Futures Trading Commission-registered exchange such as Kalshi or seek its own designation is not yet clear. The company is described as treating Arena as experimental but a top priority, and the prototype would be tested by employees on iPhone and Android before any public launch. No date has been set.

Why Meta wants in, and why it has been here before

The pull is engagement at a scale no rival can match. Meta reported roughly 3.56 billion daily active people across its app family in Q1 2026, and a prediction layer gives those users a reason to open an app repeatedly and react to live events. The category itself is compounding fast: combined trading volume on Kalshi and Polymarket reached about $50 billion in 2025 and has already passed $130 billion in 2026, according to figures cited in the reporting. Kalshi traded $33 billion in Q1 2026 alone and Polymarket $26.17 billion, and Kalshi's valuation has run from $2 billion in mid-2025 to $11 billion by December to a targeted $40 billion now. Polymarket has been raising near a $15 billion valuation. For a company with a market capitalization around $1.59 trillion, the cost of building Arena is trivial against the strategic value of owning the next engagement surface before someone else does.

The complication is that Meta has run this play before and lost. The company launched a prediction app called Forecast in 2020, which let users predict unfolding events including the early spread of Covid-19 using points rather than cash. Meta shut Forecast down in 2022, and current staff describe Arena as a rebuild of that effort. The sharper cautionary case is Diem, formerly Libra, Meta's 2019 attempt to launch a global cryptocurrency. Regulators in the United States and the European Union raised concerns over monetary sovereignty, financial stability, and privacy, and the project never shipped. The Diem Association dissolved in early 2022 and sold its assets to Silvergate Capital for $182 million. That is the outcome to weigh: when Meta moved toward regulated finance at scale, federal regulators ended it before launch. A points-only Arena sidesteps that fight for now, which is likely the point.

What it means for operators and affiliates

The market read the news as a threat to sportsbooks, not to Kalshi. On the June 23 report, DraftKings fell as much as 2% before trimming the loss to around 1%, Robinhood declined, and Flutter Entertainment, which owns FanDuel, moved lower on the day. Investors have spent the better part of a year pricing in the threat prediction markets pose to conventional sports betting operators, and a Meta entrant with billions of installed users sharpens that fear. Robinhood, which already routes event contracts from third-party venues, sits in the same blast radius.

For affiliates, the threat and the opportunity are the same fact: distribution. A points-based Arena pays no winnings and runs no licensed sportsbook, so it sits outside the affiliate revenue-share model that sends traffic to operators. But if Meta later adds real-money trading, it would do so with an acquisition funnel larger than the entire regulated US betting industry combined, and on its own terms. That is the structural worry. Affiliates compete for the top of the funnel that Meta already owns. The same legal uncertainty that has dogged the category still applies here, with state regulators and courts split over whether sports event contracts are legal betting. Federal lawmakers are circling the category too, and a points-only product launches clear of most of that fight. The harder question is what happens if Congress moves to ban or restrict prediction markets before Meta ever turns on real money, which would cap how far and how fast it could push toward cash.

What is concrete now: Meta is building Arena, it starts on virtual points, and the announcement alone knocked roughly 1% off DraftKings inside a single session.