Meta's Reported Prediction-Market Bet: Why Arena Could Pull Event Trading Into the Feed

If the reports are accurate, Meta is not simply copying Polymarket and Kalshi. It is testing whether prediction can become a mass-market engagement habit, and that should matter to every operator, exchange and regulator still arguing over where trading ends and gambling begins.

Meta has reportedly instructed a small internal team to build a standalone prediction-markets app, known internally as Arena. On the surface the product sounds cautious. It would be a mobile app, separate from Facebook and Instagram, and would use a game-style points system rather than real-money wagering at launch.

The more important question is not whether Arena ships as a regulated betting product. It is whether Meta can turn prediction into a native social behaviour before regulators, sportsbooks and exchanges have agreed what the category is supposed to be. If Meta directs even a small fraction of its audience into event-based forecasting, the competitive question shifts from "who owns prediction-market liquidity?" to "who owns the interface for speculation?"

That is a much larger fight.

What has reportedly happened

According to reporting by The New York Times, carried by Reuters and others, Mark Zuckerberg has directed a small team to develop a smartphone app similar to Polymarket and Kalshi. The app is reportedly known internally as Arena and is expected to operate independently from Facebook, Instagram, WhatsApp and Messenger. It is described internally as both experimental and a priority, and it is reported to sit alongside other standalone products in development, including an AI media tool reported as Meta Photos. The first version would probably use points rather than cash, though the reporting states that Meta has not ruled out real-money wagering later. Meta has not confirmed the product and declined to comment, and Reuters said it could not independently verify the report. This is not yet a launch story. It is a strategic signal, and it deserves to be read as one.

It is also not Meta's first attempt. In 2020 the company launched Forecast, a points-based app that invited users to predict real-world events, including the early spread of Covid-19, and it closed the product in 2022. The relevant lesson is not that the points model failed. It is that the market around it has changed completely. On Pew Research figures, combined monthly trading volume across Kalshi and Polymarket rose from under 5 billion dollars in September 2025 to around 24 billion dollars by April 2026 (these are prediction-market trading volumes, not directly comparable to sportsbook handle), and Reuters cited Bernstein's April 2026 estimate that the category could reach 1 trillion dollars in annual trading volume by the end of the decade. Forecast launched into a curiosity. Arena would launch into a market.

That market reacted quickly to the report. Within hours, investors appeared to read Meta's interest as a competitive warning. DraftKings traded lower, Flutter Entertainment (the owner of FanDuel) moved down intraday, and Robinhood, which has moved into event contracts, also came under pressure. That reaction should not be over-read as proof of Arena's eventual impact, and DraftKings still outperformed the broader market that day, but it shows how quickly investors now connect prediction-market distribution with sportsbook and brokerage valuations.

Why this matters to gambling, not just social media

Prediction markets have been pulled between three competing identities.

To their advocates they are information markets, a way to price the probability of real-world events more efficiently than polls or punditry. To parts of the financial industry they are event contracts, tradable instruments closer to derivatives than to gambling. To gambling regulators, consumer-protection bodies and many operators, they look familiar: users staking value on uncertain future outcomes.

Meta entering that argument, even experimentally, changes the scale of the question. Polymarket and Kalshi have built cultural relevance, financial-market relevance and regulatory controversy, but they remain specialist products next to Meta's distribution. Meta reported 3.56 billion family daily active people, on average, in March 2026. Arena does not need to become Facebook-sized to matter. It only needs prediction to become another lightweight behaviour that can be surfaced across feeds, creators, groups and messaging.

This is not a hypothetical contest. The incumbents have already moved. DraftKings launched its Predictions product in December 2025 and has now gone further, launching its own prediction-markets exchange, DKeX, built on the CFTC licence from its Railbird Technologies acquisition and integrated into its unified Sports and Casino app. Flutter-owned Betfair has been testing a prediction-market-style interface on its exchange. What Arena threatens is not the existence of prediction products but ownership of the front door to them.

Points are not a neutral design choice

Arena's reported points model is the part that will make the product feel softer and less regulated. It also makes commercial sense.

A points system lets Meta test three things before confronting the full weight of gambling or financial regulation: whether users enjoy forecasting real events, whether leaderboards and creator-led markets drive repeat engagement, and whether prediction mechanics generate useful behavioural data. This is a wedge strategy. Start with play-money engagement, build the habit, then decide later whether the model becomes advertising, subscriptions, sponsored markets, data products, real-money trading, or partnerships with regulated operators.

For gambling executives, the lesson from free-to-play sports games and social casino is familiar. The absence of cash staking does not mean the absence of gambling-like mechanics. It means the product is operating one step earlier in the funnel.

The question for Meta is whether Arena can make prediction feel like participation rather than betting. The question for regulators is whether that distinction survives once the product is scaled, monetised and potentially linked to real-world value.

The regulatory problem Meta cannot avoid forever

The United States is the main battleground, because prediction markets have found a path through financial-market infrastructure rather than gambling law. Kalshi operates as a federally regulated venue, Polymarket has been building a regulated US route, and brokers including Robinhood and Interactive Brokers have moved into event contracts.

That path does not remove the legal ambiguity. It relocates it. On 10 June 2026 the CFTC published a proposed rule, not a finished regime, that would amend its event-contract regulations to set out how it decides whether a contract "involves" an enumerated activity (unlawful conduct, terrorism, assassination, war or gaming) and, if so, whether it is contrary to the public interest. The proposal defines "gaming" and sets out factors distinguishing the sports-related contracts the Commission would allow from those it would treat as contrary to the public interest, with contracts resembling games of random chance among those likely to fail the test. Reporting on the proposal indicates it would bear most heavily on contracts tied to individual player actions and officiating rather than on broad sporting outcomes. Comments are due by 27 July 2026, and the framework remains a proposal rather than settled law.

Running underneath that rulemaking is a federal pre-emption fight that matters more than any single app. In April 2026, in KalshiEX v Flaherty, a divided Third Circuit panel affirmed a preliminary injunction preventing New Jersey from enforcing its gambling laws against Kalshi's sports event contracts, holding that those contracts are swaps under the Commodity Exchange Act and fall within the CFTC's exclusive jurisdiction. It was the first federal appellate ruling on the question and a significant win for the federal event-contract model, but it was a preliminary decision rather than a final merits judgment, and similar cases are still moving through other circuits. The unresolved question is how far a federal financial framework can displace state gambling law. That is the legal ground any large entrant, Meta included, would eventually stand on.

In Great Britain the position is more settled and more direct. In a February 2026 blog post, the Gambling Commission's Director of Strategy set out that commercial prediction products meeting the UK definition of gambling would need to be licensed, and that current products would most likely fall within the "betting intermediary" definition that already covers betting exchanges. In plain terms, a US-style prediction market cannot arrive in Britain and declare itself non-gambling on the basis of different terminology.

The British market also complicates Meta's options in a way the US does not. The exchange model has existed here since 2000, and Matchbook launched a UK prediction product in January 2026 under an existing exchange licence. The Commission itself questioned whether the commercial drivers behind US prediction markets would even translate to Great Britain, given how mature the domestic betting market already is. A points-only Arena might sidestep the hardest licensing questions at launch. A real-money Arena would not, and it would arrive into a market that is already served and already regulated.

Why Meta might still want the risk

Meta's interest is rational. Its advertising business remains enormous, but the company is under pressure to find growth surfaces beyond the mature social graph, with AI absorbing vast capital and the core social graph now mature. Meta's history is one of absorbing emerging behaviours from elsewhere: Stories from Snapchat, Reels from TikTok, marketplace mechanics from classifieds.

Prediction markets fit that pattern because they are content, community and commerce at once. Every market creates a debate. Every price move creates a reason to return. Every controversial event creates shareable material. Every match, election, product launch, trial or public dispute can be turned into a market-like object with a scoreboard attached.

For a social platform, that is a structured opinion format that generates its own engagement. For a gambling operator, it is unsettling. Sportsbooks already understand odds as content. Prediction markets extend that logic beyond sport into everything else.

Arena would not compete with sportsbooks in the obvious way

The lazy reading is that Meta wants to become a sportsbook. That is probably too crude.

The first real threat is not direct cannibalisation of regulated sports betting revenue. It is attention. If Arena works, it trains users to interact with odds, probabilities and outcomes outside the sportsbook environment, and it may do so most effectively with younger users who already consume markets, memes, creators and news as a single feed.

The second threat is data. A social prediction product would generate valuable signals about belief, sentiment, risk appetite and community behaviour, and Meta could combine that with one of the most sophisticated advertising and recommendation systems in existence. Sportsbooks have betting data. Exchanges have trading data. Meta would have prediction behaviour at population scale.

The third threat is partnership leverage. Meta may never need to run a regulated back end. It could own the front-end behaviour and partner with licensed exchanges, brokers or operators wherever regulation requires it, mirroring how large consumer platforms tend to monetise regulated verticals indirectly rather than owning the full compliance stack.

The comparison that matters

Model	User behaviour	Commercial engine	Regulatory exposure	Why it matters
Sportsbook	Bet on sports outcomes	Margin, trading, bonuses, media	Gambling regulation	Mature, licensed, heavily scrutinised
Betting exchange	Users bet against each other	Commission	Gambling regulation in markets such as GB	Closest gambling analogue to prediction markets
Prediction market	Trade event contracts	Fees, liquidity, market data	Split between financial and gambling regimes	Fast-growing category with unresolved boundaries
Points-based prediction app	Forecast outcomes without cash staking	Engagement, ads, data, possible conversion	Lower at launch, but not zero	Could normalise prediction at mass scale
Social prediction layer	Predictions embedded in feeds and communities	Attention, creator tools, data, sponsorship	Depends on design and monetisation	The real Meta risk to incumbents

The strategic point is simple. Meta does not need to win the regulated back end to reshape the consumer front end.

The integrity problem will follow the product

Prediction markets inherit the integrity problems of both gambling and finance, and that is no longer hypothetical. In 2026, US prosecutors brought two criminal cases alleging that individuals used non-public information to profit from Polymarket trades, one involving confidential corporate data and the other classified government information. Both remain allegations before the courts, but the direction of travel is clear: regulators and prosecutors now treat prediction-market abuse as enforceable conduct, not a grey area.

The risk grows as markets move closer to identifiable people and specific actions. A market on a final score is one thing. A market on a single player action, a referee decision, a company announcement or a political event is far more exposed to informational asymmetry. That is the kind of distinction the CFTC's proposal is trying to codify, treating contracts on narrow individual outcomes more cautiously than those on broad results.

Meta brings its own history with misinformation, coordinated behaviour, youth safety and platform manipulation. If prediction markets become social objects, the moderation problem expands. The hardest markets may not be sport but politics, where prediction mechanics would collide with Meta's existing problems around election integrity, misinformation and coordinated behaviour. Arena would need not only market rules but content rules, identity controls, age gating, manipulation monitoring and a clear view on which events should never become markets at all. Points reduce financial harm. They do not remove reputational, behavioural or integrity risk.

The supplier layer is where this gets real

For suppliers, the opportunity is not only market creation. If prediction becomes a mass-market behaviour, the required stack starts to look familiar: identity, age verification, geolocation, market surveillance, event settlement, fraud monitoring, responsible-use tooling, payments, CRM segmentation and content moderation. The winners may not be the companies with the loudest prediction-market brands, but the vendors that can make these products compliant, auditable and locally configurable.

DraftKings buying a CFTC licence and exchange technology to bring DKeX in-house is a signal in itself. When the front end scales, control of the regulated infrastructure underneath becomes the contested asset, and that is precisely the layer where suppliers sell.

What operators and suppliers should watch next

For operators, the immediate response should not be panic. Arena may never launch publicly, may remain points-only, or may fail as Meta experiments sometimes do. But the sector should track five signals.

First, distribution. Whether Meta keeps Arena standalone, or starts surfacing it through Instagram, Facebook, WhatsApp or Threads. That choice is the strategic asset.

Second, scope. Whether markets stay broad and low-stakes, or move into sport, politics and finance where the regulatory questions sharpen.

Third, value. Whether Meta introduces prizes, cash-out equivalents, tokenised value, creator monetisation or sponsorship, any of which would push the product toward gambling or financial regulation.

Fourth, partners. Whether Meta seeks licensed infrastructure rather than building the compliance stack itself, which would tell you how it intends to monetise without owning the regulated risk.

Fifth, mechanics. Whether the product starts to resemble free-to-play betting, fantasy, social casino or exchange trading. The label will matter less than the design.

The bigger story: prediction as interface

The Arena report lands as prediction markets move from niche to mainstream. Brokers are running event contracts, crypto-native venues are seeking regulated routes, sportsbooks are watching uneasily, and regulators on both sides of the Atlantic are still deciding whether these products are finance, gambling, media, or something awkwardly in between.

Meta's reported interest gives the trend a new dimension. It suggests prediction may become not only a market structure but a user interface for online debate. That should concern incumbents more than any single launch.

Sportsbooks compete on odds, product, brand, bonuses, payments, content and retention. Prediction markets compete on liquidity, breadth and probability. Social platforms compete on attention. Arena would sit at the intersection of all three.

If Meta stays with points, Arena may be dismissed as another experiment. If it adds money, partners with regulated infrastructure, or simply succeeds in making forecasting a social habit, the gambling sector will have to confront a more uncomfortable reality.

The next betting interface may not look like a sportsbook.

It may look like a feed.

---

Sources

Arena report: Reuters, "Mark Zuckerberg directed Meta to create a prediction markets app, NYT reports" (23 June 2026): https://www.reuters.com/business/mark-zuckerberg-directed-meta-create-prediction-markets-app-nyt-reports-2026-06-23/ . Underlying report: The New York Times (23 June 2026). The same Reuters report carries Bernstein's 1 trillion dollar estimate.

Forecast (2020 to 2022): Meta's New Product Experimentation team launched the points-based app in June 2020 and closed it in 2022. Quartz: https://qz.com/2069284/facebook-is-shutting-down-its-experimental-app-forecast

Trading-volume growth: Pew Research Center, "Trading volume on prediction markets has soared in recent months" (27 May 2026): https://www.pewresearch.org/short-reads/2026/05/27/trading-volume-on-prediction-markets-has-soared-in-recent-months/ . Pew's figure is notional taker volume (each contract counted at its 1 dollar notional value), so it is not directly comparable to sportsbook handle.

Share reaction (23 June 2026): TheStreet, reporting DraftKings closing down about 2 per cent, Flutter down intraday before closing slightly up, and Robinhood lower: https://www.thestreet.com/latest-news/mark-zuckerberg-makes-a-move-on-a-new-billion-dollar-market-prediction-market-arena . MarketWatch noted DraftKings still outperformed the Nasdaq that day.

Meta scale: Meta first-quarter 2026 results, "DAP was 3.56 billion on average for March 2026" (29 April 2026): https://investor.atmeta.com/investor-news/press-release-details/2026/Meta-Reports-First-Quarter-2026-Results/default.aspx

Incumbent moves: DraftKings Predictions launch, December 2025: https://www.draftkings.com/draftkings-debuts-predictions-app-entering-prediction-markets ; DKeX proprietary exchange launch, 26 June 2026 (DraftKings via BusinessWire), built on the CFTC licence from the Railbird Technologies acquisition: https://www.businesswire.com/news/home/20260624009284/en/DraftKings-Launches-Proprietary-Exchange-to-Bolster-Differentiated-Predictions-Experience ; Betfair Predicts UK beta: SBC News (April 2026).

CFTC proposed rule: Federal Register, "Prediction Markets; Public Interest Determinations", Notice of Proposed Rulemaking, 91 FR 35806 (12 June 2026), RIN 3038-AF65: https://www.federalregister.gov/documents/2026/06/12/2026-11854/prediction-markets-public-interest-determinations ; CFTC press release 9249-26 (10 June 2026): https://www.cftc.gov/PressRoom/PressReleases/9249-26 . Comments due 27 July 2026.

Federal pre-emption: KalshiEX LLC v Flaherty, No. 25-1922 (3d Cir., 6 April 2026), affirming a preliminary injunction: https://law.justia.com/cases/federal/appellate-courts/ca3/25-1922/25-1922-2026-04-06.html . More circuit rulings pending (Fourth, Sixth, Ninth).

UK position: UK Gambling Commission blog, "Prediction markets, here's what you need to know" (4 February 2026): https://www.gamblingcommission.gov.uk/blog/post/prediction-markets-heres-what-you-need-to-know

Integrity cases (both at charging stage): US Department of Justice, Southern District of New York, "Google Employee Charged With Insider Trading": https://www.justice.gov/usao-sdny/pr/google-employee-charged-insider-trading ; US Department of Justice, Office of Public Affairs, "U.S. Soldier Charged With Using Classified Information To Profit From Prediction Market Bets": https://www.justice.gov/opa/pr/us-soldier-charged-using-classified-information-profit-prediction-market-bets