Meta follows the trend to enter the prediction market. Can it avoid the same path of failure as the metaverse?

Written by: Gino Matos

Compiled by: Luffy, Foresight News

TL;DR

According to the New York Times, Meta has assembled a small team to develop an internal-codename “Arena” points-based prediction application, allowing users to bet on the outcomes of politics, sports, and global current events.

Prediction markets have already demonstrated real demand. With 3.56 billion daily active users, Meta is poised to bring a niche category to the mainstream market.

But Meta’s trust crisis, combined with scrutiny of elections and false information, could make Arena a regulatory target before it scales up.

On June 23, the New York Times reported that Mark Zuckerberg led the formation of a dedicated task force to develop the prediction market application Arena. Users will place bets on the outcomes of political elections, sports events, and international affairs using platform points.

This company—which was once renamed for the metaverse and whose Reality Labs has accumulated nearly $90 billion in losses—now turns to building a prediction market. This space has strong real demand and a user base already taking shape, but the regulatory rules are intricate and complex. This pivot may be Meta’s wisest strategic adjustment—or it could repeat previous costly failures.

The massive bill left by the metaverse

In October 2021, Facebook officially changed its name to Meta. Zuckerberg posted that the company’s core goal was to “build the metaverse,” and predicted that within a decade the metaverse would reach one billion users.

Reality Labs, the department carrying this vision, continued to expand its losses: operating losses of $17.7 billion in 2024 and $19.2 billion in 2025, bringing cumulative losses to nearly $90 billion. Meta told investors that the loss level for this segment in 2026 may be similar to 2025.

Its flagship social VR platform, Horizon Worlds, saw monthly active users fall below 200,000 in 2022, far short of the initial goal of 500,000. Meta later lowered its expectations again and plans to gradually shut down the VR version’s operations in 2026.

Why prediction markets are a completely different track

In 2026, the two leading platforms, Kalshi and Polymarket, are expected to have a combined monthly trading volume of about $24 billion. Industry institutions predict that the sector’s annual prediction market trading volume will exceed $130 billion.

Robinhood launched a prediction market section in 2025, and Interactive Brokers also integrated event contracts into its trading platform. Even the Golden Globe Awards has introduced interactive prediction market segments. A Bernstein research report in April estimated that the sector’s annual trading volume could reach $1 trillion by 2030.

Meta has always been good at copying popular products and using its massive traffic to overtake competitors: after Snapchat launched Stories, Instagram launched Stories; after Twitter dominated the social text-and-image track for a decade, Meta launched Threads; after TikTok went viral in short video, Meta launched Reels. As of April, Meta’s entire suite of products had 3.56 billion daily active users—dwarfing the traffic volumes of all existing prediction market platforms.

Arena uses a points-based design, continuing Meta’s usual strategy: capture users’ existing behavioral demand, embed them into Meta’s own traffic ecosystem, and rely on large-scale distribution to make up for shortcomings in original product design.

Building a prediction market requires only software, information feeds, account systems, content moderation, and compliance systems, and in some scenarios it can connect with licensed partner institutions. But the metaverse needs custom hardware, immersive content, virtual avatars, and dedicated operating environments—and it takes years to cultivate user habits. Reality Labs’ massive losses show that creating an entirely new track model from scratch is extremely costly.

A comparison of the core dimensions between the metaverse and prediction market Arena

Arena is not Meta’s first attempt at prediction markets; the previous product was already shut down

As early as the beginning of the COVID-19 pandemic in 2020, Meta launched a points-based public prediction app called Forecast, focusing on predictions of current events, but it was shut down in 2022. At that time, Polymarket had not yet become popular following the 2024 U.S. presidential election, and Kalshi had not won in its lawsuit with the U.S. Commodity Futures Trading Commission (CFTC) over election contracts. The industry’s annual trading volume also had not yet exceeded $50 billion.

There are no shortage of regulatory penalty cases in the track Meta is about to enter:

In 2022, the CFTC ruled that Polymarket was conducting unregistered off-exchange event derivatives trading and imposed a $1.4 million fine;

Kalshi spent years in federal litigation fighting over eligibility to operate election contract business. In September 2024, a district court issued a favorable ruling; in May 2025, the CFTC withdrew its appeal. Although compliance space for election event contracts has opened up, controversies over political trading and market fairness have not fully subsided;

In April 2026, the CFTC initiated what it said would be the first-ever insider trading lawsuit in the history of prediction markets, accusing an active-duty U.S. military officer of using confidential intelligence about a Venezuelan operation to profit from trades on Polymarket.

Meta’s past financial product layouts have already made regulators highly wary of its financial ambitions. Facebook-led digital stablecoin project Diem (formerly Libra) was ultimately sold at a low price to Silvergate Bank in 2022 after regulators determined that Meta’s control of a payment network serving billions of users would result in excessive concentration of financial and social power. During that year’s Libra hearings, Meta’s combination of social identity, political content, financial incentives, and market data faced strong opposition from regulators.

Precisely because points-based prediction games can bypass strict early financial regulation, Meta chose this approach as Arena’s starting point.

What advantages can come from massive traffic

The most feasible form for Arena’s initial product is to build a mass prediction feature by leveraging social scale: Instagram creators publish prediction markets for award events; Facebook groups discuss sports odds; WhatsApp communities share collective prediction views; and Meta AI aggregates mainstream expectations across the internet.

For now, this version does not involve cash event contracts, which have previously drawn regulatory penalties. It relies solely on the 3.56 billion daily active users’ social graph to operate.

However, the core logic of prediction markets is to constrain prediction behavior and form fair prices through real-money market competition. Once it is replaced with points-based interactive incentive mechanisms, the product will prioritize virality and user time spent rather than accuracy of prediction outcomes.

Meta’s poor track record in handling political content and cracking down on misinformation means regulators and the media will naturally view every controversy Arena generates with scrutiny.

Meta’s traffic advantage is enough to support the sector’s scale. The success logic of Stories and Reels is consistent: capture users’ existing preferences and amplify their distribution through a platform with billions of users. If Arena builds lightweight social prediction features, controls the financial capital threshold, and enables ordinary Facebook users to easily access prediction markets, while platforms such as Kalshi maintain a professional trading positioning, Meta may expand the industry’s “pie,” benefiting existing leading platforms.

Crypto-native users with financial understanding can support a trillion-dollar prediction market track, and Meta’s 3.56 billion daily active users represent an enormous pool of ordinary users that the industry has never reached—making this the biggest opportunity for this entry.

But just two months before news of Meta’s entry was exposed, the CFTC had just launched the first insider trading lawsuit in prediction market history, and regulatory scrutiny of the industry continued to tighten. Meta’s platforms cover prediction markets related to elections, sports events, and predictions involving public figures, making it highly likely to trigger regulatory intervention. Combined with the company’s negative record in handling sensitive political content, Meta enters with an inherent credibility weakness, and its massive traffic may in turn amplify all kinds of negative controversies.

Four development outlooks for Arena

Many of Meta’s previous financial products failed completely because regulators deemed the trust issues unsolvable.

Arena has inherent advantages: the prediction market track has already taken shape, and real existing users are there. But Meta, as the operating platform, carries the same negative reputation as it did when Libra failed. Once elections and capital transactions are involved, trust is a core asset that Meta must earn through long-term operations; simply relying on the scale of traffic cannot compensate for the lack of credibility.