Zuckerberg đã để mắt đến việc mua lại Kalshi trước khi phát triển ứng dụng dự đoán riêng của Meta.

Meta chief executive Mark Zuckerberg approached prediction-market leader Kalshi about a takeover before ordering his own team to build a competing app, according to a new report.

• • Zuckerberg met Kalshi CEO Tarek Mansour about a buyout in 2025, but talks stalled and never reached a formal offer
• • Meta is now building a play-money prediction app, Arena, sidestepping the gambling classification dogging Kalshi and Polymarket
• • Kalshi raised $1B in May at a $22B valuation, double its $11B mark five months earlier

How a buy-then-build turned into a contested market

NPR reported Monday that Zuckerberg met Kalshi co-founder and CEO Tarek Mansour last year to discuss buying the exchange as its user numbers climbed, citing three people who were not authorized to speak publicly. The talks never advanced to a formal offer. Accounts of the collapse diverge: some sources said Mansour was unwilling to sell, while others said Meta judged the legal and regulatory questions around Kalshi too messy to take on.

Rather than walk away from the sector, Zuckerberg directed staff to build a standalone app, internally named Arena. The design marks a deliberate departure from the market leaders, as it will use play money rather than real-cash wagers as users guess on news events and trending topics. Reportedly, Meta’s artificial-intelligence systems will generate the questions and settling outcomes. Neither Meta nor Kalshi commented on the talks.

The real-money stakes and the associated distinction between financial and gambling products has drawn dozens of legal challenges in just the United States, and produced an even more hostile regulatory posture in Europe. Minnesota became the first state to make it a felony for such platforms to operate, and the U.S. Justice Department has opened two insider-trading cases tied to Polymarket – one involving a special-forces soldier accused of trading on classified information about the capture of Venezuelan leader Nicolás Maduro, another a Google employee accused of using confidential search-trend data. By keeping cash out of Arena, Meta aims to avoid the classification fight altogether, though it also removes the profit motive for users.

Kalshi raised $1 billion in a Series F round in May led by Coatue, with Sequoia, Andreessen Horowitz, and Paradigm participating, at a $22 billion valuation – doubling from just five months ago. Kalshi told Bloomberg its annualized revenue tops $1.5 billion, and said institutional trading had risen 800% over six months. Rival Polymarket, which runs an offshore exchange outside U.S. regulatory reach, was valued at $10.7 billion, according to Pitchbook.

Sector volumes have swung sharply through 2026. Prediction-market platforms recorded roughly $8.6 billion in taker volume in April and about $29.8 billion in notional terms, with Kalshi overtaking Polymarket for the monthly lead, according to Dune Analytics data. Monthly figures peaked near $25.7 billion in March, and the total 2025 industry volume topped $63 billion.

Last year, the Federal Trade Commission (FTC) argued at trial last year that Meta runs a “buy or bury” strategy, either acquiring young rivals or cloning them to squeeze them out. A judge sided with Zuckerberg’s company, finding it broke no competition law in acquiring Instagram and WhatsApp; the FTC is appealing. Meta also struck a partnership with Kalshi in March, allowing its markets to integrate with the Threads app. Now, it seems the calculus moved from “buy” to “bury.”