Meta’s New Prediction Market App ‘Arena’ Takes Aim at Kalshi and Polym

Mark Zuckerberg has reportedly told his teams to build a Meta prediction market app — and if it actually ships, platforms like Kalshi and Polymarket should be paying close attention. According to a report from The New York Times, citing two employees with direct knowledge of the project, the app is internally called ‘Arena’ and is being treated as a top priority inside the company.

• Meta prediction market app ‘Arena’ is being developed at Zuckerberg’s direct order, using points instead of real money.
• The Meta prediction market will launch as a standalone app, separate from Facebook and Instagram, according to insiders.
• Kalshi and Polymarket could face serious competition from Meta’s 3.56 billion daily active users if Arena launches publicly.
• Regulatory scrutiny over prediction markets is intensifying in the US, complicating the landscape for any new entrant.

What We Know About Arena So Far

The Meta prediction market app would let users place wagers on future events — elections, sports outcomes, cultural moments — but crucially, using a points system rather than real money. That distinction matters enormously, both commercially and legally. By avoiding direct financial stakes, Meta can sidestep the financial services regulations that have made real-money prediction platforms a regulatory minefield in the United States.

Arena is also reportedly being built as a standalone mobile app, entirely separate from Facebook, Instagram, WhatsApp, or Threads. That’s an interesting strategic choice. Meta has a habit of folding new features into its existing apps — Reels got absorbed into Instagram, Marketplace lives inside Facebook — so building Arena as its own product suggests the company sees a genuinely distinct audience or use case here, not just a bolt-on engagement feature. As a standalone Meta prediction market, Arena would need to build its own habit loop from scratch rather than borrowing one from an existing platform.

Insiders described the effort to the Times as ‘experimental,’ which is the kind of language companies use to manage expectations before a launch. But they also called it a top priority — and in corporate-speak, those two things rarely sit comfortably together unless serious resources are being committed. The fact that this Meta prediction market effort is being framed both ways suggests leadership is keeping its options open while still pushing the project forward at speed.

Why the Meta Prediction Market Could Shake Up the Industry

To understand the potential scale here, consider that Meta reported 3.56 billion daily active users across its family of apps as of March. Kalshi and Polymarket are legitimate platforms with real traction — Polymarket in particular became something of a cultural moment during the 2024 US presidential election cycle, with its odds being cited by journalists and pundits alike — but their total user bases are a rounding error compared to Meta’s reach.

Distribution is almost always the decisive factor in consumer internet. A Meta prediction market app that can be cross-promoted to even a fraction of Meta’s audience, with a frictionless sign-up because users already have Meta accounts, has a structural advantage that no startup can easily replicate. The points-based model also removes the hesitation many casual users feel about putting real money on the line, which could open the Meta prediction market to a far broader demographic than existing platforms currently serve.

Kalshi, for its part, fought a lengthy legal battle with the Commodity Futures Trading Commission to operate its real-money contracts in the US, eventually winning the right to list political event contracts. That regulatory moat is real, but it’s not the kind of moat that stops Meta from building a points-based alternative that attracts the same eyeballs without the same regulatory burden.

Meta’s Complicated History With Financial Products

This wouldn’t be Meta’s first attempt to build something at the intersection of social media and money. Back in 2019, the company unveiled Libra — a stablecoin project backed by a consortium of major companies that briefly looked like it might reshape global payments. The backlash from regulators and lawmakers was swift and severe. The project was rebranded to Diem, stripped down, and quietly killed off in 2022.

More recently, Meta has been more cautious in its financial product experiments. In April, the company rolled out USDC payouts for select Facebook creators in Colombia and the Philippines — a narrow, geographically contained test that kept it well clear of US financial regulators. Still, US lawmakers have already flagged concerns about Meta’s longer-term stablecoin ambitions on home soil.

The points-based structure of Arena reads partly as a lesson learned from Libra. By keeping real money out of the equation entirely, the Meta prediction market avoids triggering the securities and commodities frameworks that have made real-money prediction markets so legally contested. It’s a smart move — if slightly less exciting for users who enjoyed the genuine financial stakes on Polymarket.

The Regulatory Storm Surrounding Prediction Markets

Arena arrives at a moment when prediction markets are under more scrutiny than ever. The CFTC is currently engaged in legal disputes with several state regulators over who has authority to oversee these platforms, and Congress is actively considering legislation aimed at closing loopholes that allegedly allow people with non-public information — think government officials or military personnel — to profit from event contracts.

That last concern was brought into sharp focus by the case of Gannon Ken Van Dyke, a US soldier accused of making more than $400,000 on a Polymarket contract tied to the capture of Venezuelan President Nicolás Maduro, who was removed by US forces in January 2025 and is now facing a criminal trial in New York. Van Dyke is scheduled to go to trial in December. The case has become a reference point for lawmakers arguing that the current regulatory framework is inadequate.

For Meta, entering this space — even with a points-only product — means walking into a political environment where prediction markets are actively associated with insider trading concerns and national security implications. Any Meta prediction market launch, however carefully structured, will attract the attention of legislators who are already primed to scrutinize the category. That’s not insurmountable, but it does suggest Arena will face scrutiny even if it never touches real money.

Meta Prediction Market Ambitions in the Broader AI Pivot

There’s another layer worth considering. Meta is in the middle of a significant strategic pivot toward artificial intelligence. The company reportedly cut around 10% of its workforce in April — roughly 8,000 people — as part of that shift. AI is clearly where the executive team is putting its biggest bets, and several of Meta’s recent product moves, including the expansion of its Meta AI assistant, reflect that priority.

Where does a Meta prediction market fit into an AI-first strategy? Potentially quite well. Prediction markets generate rich, structured data about collective human expectations — data that could feed directly into AI systems that model future events. There’s also a natural synergy with Meta AI as a product: imagine an assistant that can surface the current consensus odds on any event a user asks about, drawn from Arena’s active markets. That’s a differentiated feature no standalone prediction market can easily replicate.

Whether Arena ever ships — and in what form — remains genuinely unclear. The ‘experimental but top priority’ framing gives Meta maximum flexibility to pull the plug if regulatory headwinds intensify or the product doesn’t test well. But the fact that Zuckerberg is reportedly directing this work personally suggests it’s more than a skunkworks side project. For Kalshi, Polymarket, and anyone else building in the prediction market space, a Meta prediction market entering the category is a serious signal worth acting on now rather than later.

Source: Cointelegraph

Frequently Asked Questions