The Wall Street Journal published an investigation on June 21 revealing that Polymarket, the world’s largest prediction market, paid online creators to produce deceptive videos simulating lucrative bets on “near-perfect copies” of its website. The clips featured trades and winnings that were not real, were amplified by a paid “social-media army,” and reached more than 140 million views before the story broke.
The finding is not simply a marketing scandal. It is a proof-of-concept for how prediction markets — platforms that claim to aggregate collective wisdom into reliable probability estimates — can be gamed at every level, from the data they run on to the public perception that drives their liquidity.
The WSJ analyzed 1,105 videos posted by 10 creators paid by Polymarket between December 2025 and mid-May 2026. Roughly 70% of the videos displayed a bet, and none of the approximately US$1.9 million (about £1.5 million) in wagers shown were executed on Polymarket’s actual platform. Many clips were filmed on imitation domains such as poiymarket.com, designed to resemble the real product.
In 118 videos, creators celebrated roughly US$900,000 (about £710,000) in fabricated wins. The WSJ reported that the same bets would have lost more than US$166,000 (about £131,000) if placed on the actual market. One video showed a creator claiming a US$100,000 win after Donald Trump allegedly said “McDonald’s” in January; the WSJ said Trump did not say it publicly that month, and the clip used was older. Public market data indicated more than 50 real accounts took the referenced bet, and all lost.
Creators were paid about US$2,000 to US$3,000 (about £1,580 to £2,370) per month and were instructed not to disclose the payments. A marketing contractor then amplified the clips, reaching more than 140 million views.
Razeen Khan, a college student and creator who worked with Polymarket until March, compared the practice to commercials that make fast food look better than it is. “We’re depicting what actually happens,” he told TechCrunch.
Polymarket said it is “committed to maintaining accurate, fair, and transparent markets” and plans to conduct an audit of its promotional content.
The machinery of manufactured belief
This scandal is dangerous precisely because prediction markets sell themselves as a cure for misinformation. The pitch is elegant: put money behind your beliefs, and the market price reflects the collective probability of an event. Polymarket’s defenders argue that genuine trades settle on the Polygon blockchain using USDC stablecoins, that outcomes are resolved by UMA’s permissionless oracle system, and that every position is publicly auditable.
None of that matters if the public narrative is built on fake data that drives real trading volume. The distinction between “the market is real” and “what people believe the market is saying is fake” is a distinction without a difference when the two cannot be disentangled from the outside.
The same mechanism that makes prediction markets powerful — that belief drives price, and price reinforces belief — makes them vulnerable to capture. A coordinated campaign fabricating wins, implying insider knowledge, and simulating easy money does not just attract naive users. It signals to everyone that the market is a slot machine rather than an information aggregation engine. And when the public cannot tell which surface-level signals are real, the informational value of the entire system collapses.
Beyond Polymarket: why this matters for every system built on trust
The stakes of the Polymarket scandal extend far beyond prediction markets. The technique on display — pay creators to fabricate compelling narratives that look authentic, amplify them to millions, claim the resulting attention as validation — is the same method that has been used to manipulate stock prices through fake news, to distort sports betting lines through fabricated injury reports, and to sway political opinion through coordinated disinformation campaigns.
What makes this cycle so corrosive is its self-reinforcing nature. Manufactured narratives drive real economic activity. Real economic activity is then cited as proof that the narratives were accurate. The people who profit from the gap between perception and reality are almost always those who can afford to manufacture perception in the first place: the already ultra-wealthy, operating through well-funded marketing operations.
Polymarket’s own trajectory illustrates the pattern. The company was fined US$1.4 million by US regulators in 2022 for operating an unregistered market, later reincorporated in Panama, and has since secured a path to enter regulated US markets through its 2025 acquisition of licensed exchange QCEX. Throughout this journey, the company’s valuation has risen, buoyed by the viral reach of content that the WSJ has now shown to be manufactured.
The regulatory vacuum
The timing of the scandal is particularly awkward for prediction markets. Minnesota became the first US state to ban them in May 2026. Spain blocked both Polymarket and rival Kalshi the same month as it investigates whether the platforms violate gambling law. The US Commodity Futures Trading Commission, which has been scrutinising Kalshi’s affiliate model for distortion of election-contract prices, now faces pressure to treat Polymarket’s paid-influencer program as a market-integrity risk.
But regulation alone cannot solve the deeper problem. The Polymarket scandal is not a case of bad actors breaking clear rules. It is a case of a platform leveraging the normal tools of social-media marketing to create a false impression of the normal functioning of its product. The behavior is deceptive. Whether it is illegal depends on jurisdictions that have not yet decided what prediction markets are: gambling, securities trading, or something new that existing laws do not cover.
In that ambiguity, platforms have every incentive to push the boundaries. And every time one gets caught, the pattern it reveals — fake platforms, paid influencers, manufactured virality — becomes a template for the next.
Sources: TechCrunch (June 21); The Verge (June 21); Engadget (June 21); BingX (June 21); Prediction News (June 20)