Prediction markets, which have rapidly gained popularity, have faced a serious challenge. A large-scale investigation has uncovered approximately 34,000 suspicious transactions on the Polymarket platform, totaling around $200 million between August 2025 and June 2026. This is not just a statistic—it is a signal of a systemic problem threatening the very essence of decentralized predictions.
Polysights, a service specializing in blockchain data analysis, evaluated each transaction based on eight key criteria: bet size, account age, event probability at the time of entry, and volume concentration. This approach helps identify anomalies but does not prove the presence of insider trading. Some participants may have simply been better prepared or taken a successful risk. However, the numbers speak for themselves.
Anomalies and Profit Concentration
The peak of suspicious activity occurred in late February, coinciding with geopolitical events surrounding Iran. Profit distribution was extremely uneven: more than half of all winnings went to just 1% of the most successful wallets. Moreover, 57% of these wallets were created less than a day before the transactions were made. This suggests the deliberate use of non-public information.
Particular attention should be paid to a group of 38 linked addresses. These traders bet on U.S. actions in Iran and Venezuela with 98% accuracy, earning $1.6 million and withdrawing funds through a single Coinbase account. Such statistics leave no room for chance.
Reaction from Regulators and Platforms
Authorities in the U.S. and Israel have already responded. Military member Gannon Ken Van Dyke was accused of earning over $400,000 from secret data about the operation in Venezuela. In Israel, similar charges were brought against a reservist who bet based on classified military information.
Polymarket tightened its rules in late March, banning transactions using stolen information and handing over data on nearly 100 wallets to law enforcement. Kalshi also reported suspicious activity by a White House broadcast operator to the regulator. Goldman Sachs restricted employee access to prediction markets, leaving only sports and entertainment bets.
However, there is an alternative viewpoint. Economist Robin Hanson argues that insiders make prices more accurate, and secret-keepers should remember that traders always have an incentive to uncover them. This is hard to argue with, but for the young prediction market industry, reputational risks currently outweigh the potential benefits of "smart" bets.
Expert opinion: Prediction markets are a powerful tool for aggregating information, but their transparency is becoming an Achilles' heel. Platforms will have to find a balance between decentralization and protection against manipulation, otherwise regulatory pressure could stifle innovation in its infancy.