Polymarket Refers At Least 2 Midterm Bets to CFTC After Insider-Trading Probes

Two Cases Referred to the CFTC

Polymarket has investigated dozens of bets on its U.S. midterm election markets for possible insider trading and manipulation, and at least two of those cases have been referred to the Commodity Futures Trading Commission (CFTC). The enforcement push sits with Shana Bautista, the company's head of global intelligence and investigations, a former FBI official who joined Polymarket in June and runs a 16-person team policing election bets with a new surveillance system ahead of the November 3 vote. The platform pulls social media activity, Federal Election Commission filings, records from crypto firms, banking partners and statistical models into a profile of every user, and investigators receive hundreds of alerts a day when an account trips enough internal red flags. Bautista declined to disclose the tripwire criteria, citing system security. Her message to would-be insiders is blunt: “If you have information that you have a duty to protect, don't trade on it. If you decide that you are smarter and can do it, we will find you.” The team covers both Polymarket's identity-verified U.S. exchange and Polymarket International, where a user needs only a crypto wallet to trade anonymously. Even there, she said, anonymity rarely holds: “There's never been a case that we work where we haven't been able to identify, potentially, who these people are.”

Democrats at 92% for the House

The probes land as election betting absorbs record volume. Polymarket runs 506 election markets, and its House control contract, with $24 million traded, prices a Democratic takeover at 92% against 9% for Republicans holding the chamber. The $9 million Senate market gives Democrats a 66% implied probability versus 35% for the GOP. On the combined power-balance contract, a Democratic sweep of both chambers trades at 65%, a split outcome at 27%, and a Republican sweep at just 8%. Texas leads the state boards with $7 million traded, where Democrat James Talarico carries a 67% implied win rate against 34% for Republican Ken Paxton. Michigan and Maine follow at $3 million each: Abdul El-Sayed leads Mike Rogers 69% to 32%, and Troy Jackson leads Senator Susan Collins 62% to 39%. Ohio's Sherrod Brown trades at 70% against Jon Husted, while Alaska's Mary Peltola sits at 75% over Dan Sullivan, both markets near $2 million. Governor races point the same direction: Xavier Becerra in California trades at 97%, Byron Donalds in Florida at 80%, and incumbent Greg Abbott in Texas at 69%. California's proposed billionaire wealth tax has drawn $4 million in volume, with 77% of it leaning toward rejection. Across the industry, more than $1.4 billion has already changed hands on 2026 midterm markets, with rival Kalshi near $1.1 billion and Polymarket above $340 million.

Concentrated Bets, Thin Anonymity

Regulatory attention has tracked the concentration of that money. A study by Rutgers professor Harry Crane for the Coalition for Prediction Markets estimated Americans wagered as much as $34 billion on offshore prediction markets in the 12 months through April 2026, with up to 30% of Polymarket's volume in that period coming from U.S. users barred from the offshore site. The Anti-Corruption Data Collective, reviewing more than 21,000 midterm markets across Kalshi and Polymarket, found the top 1% of Polymarket wallet addresses drove 70% of congressional trading volume this cycle. On-chain data shows 24,930 Polymarket wallets held midterm positions as of October 18, with a median position of just $14, while the top 100 wallets controlled 56% of all position value. Concentration alone is not proof of wrongdoing, and on-chain records persist whether a trader keeps funds on a venue or in self-custody, a distinction we explain in Crypto Exchange vs Crypto Wallet. The surveillance flags clusters of accounts behaving alike, a pattern that can indicate collusion or copy trading as easily as outright manipulation. The concern is not theoretical: more than 80 Polymarket users placed suspiciously timed bets, including wagers before undisclosed U.S. and Israeli operations against Iran, prompting House Oversight Chairman James Comer to open an investigation in May. An active-duty Army soldier was separately charged with betting classified details of the Maduro capture operation, turning more than $33,000 into roughly $410,000; he has pleaded not guilty.

Self-Policing Before November 3

The referrals show where enforcement of election markets currently lives: inside the platforms themselves. The CFTC has largely left policing to the companies, though its staff warned in September that mention markets, contracts on whether a named person says certain words, should be presumed readily susceptible to manipulation. Chairman Michael Selig has said the agency will watch for manipulation in prediction markets ahead of the midterms, and past cases set the ceiling: a former White House teleprompter operator was ordered to return $107,539.02 plus a $65,000 penalty, and former Representative George Santos $17,569.98 plus $17,500, both with three-year trading bans. Barnard College economist Rajiv Sethi warns the odds themselves can shape fundraising and coverage, a self-fulfilling prophecy risk that grows when a few large accounts act like a market maker in size. The number to watch now is whether those two referrals become formal enforcement before the November 3 vote.