The fine is $35,000. The message is considerably more expensive.
The Commodity Futures Trading Commission has ordered former U.S. Congressman George Santos to pay $35,000 for manipulative trading in prediction markets. Not for campaign finance fraud. Not for fabricating a résumé. For trading. The same activity that millions of users execute daily on platforms like Polymarket, Kalshi, and PredictIt.
Let the record show what this actually is: the first individual-level enforcement action against a prediction market user in CFTC history. Every prior action targeted platforms. This one targets the person clicking the buttons.
The penalty is pocket change for a federal enforcement action. That is the point. The CFTC is not collecting revenue. It is establishing jurisdiction, building precedent, and serving notice that the Commodity Exchange Act applies to whoever touches an event contract, regardless of whether the venue calls itself decentralized, blockchain-based, or “just a sports prediction.”
Silence in the ledger speaks louder than hype. This ledger has just been written into the regulatory canon.
George Santos is not a typical market manipulator. He is a former member of the United States House of Representatives who pleaded guilty in August 2024 to federal charges of wire fraud and aggravated identity theft tied to campaign finance schemes. His political career collapsed amid revelations of fabricated biographical claims. The ethical and legal wreckage has been exhaustively documented.
What has received far less attention is the CFTC’s separate enforcement action against him for trading behavior in prediction markets. The commission’s order alleges manipulative trading—a broad category that, in low-liquidity event contract markets, can encompass wash trading, spoofing, and coordinated self-trading designed to move settlement prices.
To understand why this matters, you need the regulatory backdrop.
The CFTC has spent the past three years wrestling with its authority over event contracts. In 2022, it fined Polymarket $1.4 million and forced the platform to block U.S. users. In 2024, Kalshi sued the CFTC over the agency’s attempt to block congressional control markets—and won, with a federal court compelling the commission to allow the contracts. In January 2025, the CFTC issued a notice of proposed rulemaking seeking to restrict or prohibit certain event contracts, particularly those tied to political contests and other activities the agency deems contrary to the public interest.
Into this evolving legal landscape drops the Santos case. The CFTC is clearly testing its enforcement authority not just against platforms, but against users themselves. The legal theory: the Commodity Exchange Act’s anti-manipulation provisions apply to anyone engaging in manipulative or deceptive trading practices in CFTC-jurisdictional markets—and prediction markets, in the CFTC’s view, are CFTC-jurisdictional markets.
Based on my experience decoding 500+ pages of SEC filings during the 2024 ETF approval cycle, I can tell you that small-dollar enforcement actions are never about the dollar figure. They are legal position markers. They exist to establish a favorable reading of statute before the big cases arrive.
The technical question is harder than the legal one. How does one manipulate a prediction market, and why are these markets so vulnerable to it?
The answer begins with a structural flaw I have observed since my 2020 work analyzing DeFi yield protocol mechanics: low-liquidity order books are price-taker magnets. The mechanisms that make prediction markets efficient in theory—continuous price discovery, open participation, settlement by real-world outcome—are the same mechanisms that make them manipulable in practice when depth is thin.
Consider the mechanics. A typical event contract on a mid-tier prediction market might have an order book with $5,000 in resting bids on one side and $8,000 in resting offers on the other. The bid-ask spread might be two or three cents on a contract trading at $0.40. A trader with $10,000 can sweep the visible book, push the price to $0.55, and trigger a cascade of algorithmic responses from participants who interpret the move as information.
Except it is not information. It is fabrication. The trader is not expressing a view about the probability of the event. They are manufacturing a signal.
The most common technique, and the one the CFTC likely documented in the Santos case, is wash trading: executing buy and sell orders against oneself to create artificial volume and price movement. In a low-liquidity market, a small number of wash trades can shift settlement prices meaningfully. The trader can then profit from positions taken in correlated markets—a prediction market on one platform, a derivatives contract on another, or simply a reverse position taken before the manipulation campaign begins.
This is where the analysis gets interesting from a technical standpoint. The audit trail never lies, only the auditor can. In traditional markets, identifying wash trading requires sophisticated surveillance systems that correlate accounts, IP addresses, device fingerprints, and execution timing across multiple venues. The process takes months and often requires subpoenas across borders.
In blockchain-based prediction markets, the evidence collection problem inverts. Every order is on-chain. Every wallet address is visible. Every transaction history is permanent. The trading pairs, the timing, the quantity patterns—all of it is sitting in a public database, waiting for an analyst to connect the wallet to a human.
The CFTC did not need to crack encryption to build its case against Santos. It needed to trace account records, link them to a known individual, and overlay the trading pattern on the price impact. The decentralized architecture that prediction market proponents cite as a feature—transparency—became the enforcement mechanism.
Let me be precise about the manipulation taxonomy. In my audit experience, low-liquidity event markets exhibit four recurring vulnerabilities.
First, order book sparsity. When the book has insufficient resting liquidity, any market order moves price disproportionately. This is the entry-point vulnerability. A manipulator does not need to be sophisticated; they need the market to be shallow.
Second, settlement oracle dependency. Prediction markets settle against a data source—an election result, a sports outcome, an economic indicator. If the settlement source is ambiguous or delayed, there is a window for price distortion before the oracle confirms reality.
Third, cross-platform price divergence. Prediction markets do not share a consolidated tape. Polymarket shows one price for a given event; Kalshi shows another; PredictIt shows a third. These spreads create arbitrage opportunities in normal conditions and manipulation surface area in abnormal ones. A trader can push the price on a thin platform and harvest the discrepancy on a thicker one. No unified price discovery standard exists across the industry. That gap is an open invitation.
Fourth, leverage and position limits. Many prediction platforms allow leveraged positions without rigorous margin requirements. This amplifies the incentive to manipulate: a 10x leveraged position in a manipulated market produces 10x the profit for the same capital.
Now, the specific case. The CFTC’s order against Santos does not disclose the platform, the contracts traded, or the forensic methodology. This is common in enforcement actions—the commission reserves details for the litigation record. But the structural inference is straightforward.
Santos, according to the CFTC, engaged in manipulative trading. The most plausible mechanism, given the context of low-liquidity political event contracts, is self-matching or coordinated cross-account trading designed to simulate market interest. The penalty amount suggests the total illicit profit was modest—the CFTC’s civil monetary penalty framework scales with the gravity of the violation and the defendant’s ability to pay. A $35,000 fine for a disgraced politician with limited assets is consistent with a deterrence-focused action rather than a restitution-driven one.
The data does not negotiate; it only confirms. And the data here confirms that prediction markets have a liquidity problem that is also a manipulation problem.
The deeper issue is the settlement mechanism. A prediction market’s price is supposed to converge on the true probability of an event. In an efficient market, arbitrageurs correct mispricing. But arbitrage requires capital, and capital requires confidence. When traders observe a suspicious price move that is later confirmed to be manipulation, their confidence erodes. The liquidity retreats. The market becomes thinner. A thinner market invites the next manipulator.
This is the negative spiral the CFTC’s action, however small in dollar terms, attempts to interrupt.
But let me be clear about what the CFTC is not doing. It is not attacking the underlying technology. It is not questioning the legitimacy of event contracts as a product category. It is targeting conduct. That distinction matters, and it brings us to the economic analysis of the sector.
Prediction market platforms capture value through trading volume and liquidity depth. If regulatory pressure causes U.S. market makers to withdraw, the global liquidity pool shrinks. Thinner books mean easier price manipulation. Easier manipulation invites more enforcement. More enforcement accelerates withdrawal. The feedback loop is structural, not incidental. Every compliance announcement from a platform tightening KYC or restricting political contracts should be read through this lens.
The market structure question also exposes a competitive divergence. Kalshi has spent years building a compliant, court-tested U.S. franchise. Polymarket, after its 2022 settlement and subsequent U.S. block, made a calculated bet to re-enter the U.S. market and rode the 2024 election cycle to explosive volume growth. PredictIt operates under an academic-research exemption with strict position limits. Azuro and Augur occupy the decentralized end of the spectrum, with minimal U.S. exposure but also minimal liquidity relative to the centralized leaders.
This enforcement action hits each platform differently. Kalshi’s validated compliance posture becomes a moat. Polymarket faces renewed uncertainty about its U.S. strategy. The decentralized venues get a warning that transparency cuts both ways. The CFTC has demonstrated it can build a complete evidence chain—trading behavior, identity linkage, profit results—using exactly the public ledger data that the decentralized ethos celebrates.
The conventional reading of this enforcement action is bearish: the CFTC is tightening its grip on prediction markets, and the sector will be strangled by regulation.
I disagree. The contrarian read is more interesting and, I believe, more accurate.
This action signals that the CFTC is treating prediction markets as legitimate financial infrastructure subject to standard anti-manipulation rules—not as illegal gambling operations to be shuttered. There is a critical distinction between the CFTC’s enforcement posture toward Polymarket in 2022 and its posture toward Santos in 2025. In 2022, the agency fined the platform and forced it to cut off U.S. users. In 2025, the agency fined an individual for manipulating trades.
The target changed. That is the story.
By pursuing the user rather than the venue, the CFTC implicitly accepts the venue’s existence. It is saying: prediction markets can operate, but the laws against manipulation apply to you, the trader. That is a compliance framework, not a prohibition framework.
Consider the counterfactual. If the CFTC wanted to kill prediction markets, it would not bother with a $35,000 fine against a single individual. It would issue a proposed rule banning political event contracts outright, as it attempted to do with Kalshi. That rulemaking is still pending. But this enforcement action does not advance a ban agenda; it de-risks one. The CFTC can point to the Santos case and say: “We are not banning event contracts. We are punishing manipulative conduct.”
This is how financial markets mature. The first enforcement actions in any new asset class are small-dollar, symbolic cases designed to establish norms. The SEC’s early crypto enforcement followed exactly this pattern: minor actions against minor actors, building precedent for later, larger cases.
For prediction market platforms, the strategic implication is clear. The platforms that invest in surveillance infrastructure, KYC/AML compliance, and market integrity controls will thrive. The platforms that continue to operate as anonymous, lightly-regulated venues will face a shrinking user base and increasing legal risk. The market is not being killed. It is being professionalized. Yield is not income; it is risk repackaged—and in this sector, the risk repackaging is now subject to federal enforcement.
The governance tension cannot be ignored either. Decentralized platforms that rely on DAO governance will face escalating friction between community demands for open access and compliance teams demanding user restrictions. Every enforcement action shifts power toward the compliance function. That is not a bad outcome; it is the price of operating in a regulated financial system. But it should be named for what it is: a centralization pressure that decentralization purists will have to confront.
There is also a political dimension worth noting. Santos was a federally indicted, politically toxic figure when the CFTC acted. Prosecuting him carries zero political risk. It is the cheapest possible target for establishing precedent. Any competent regulatory strategist would have flagged this case as a low-cost, high-leverage opportunity. The CFTC took it.
Watch the rulemaking calendar, not the fine number. The CFTC’s final event contract rule is the tectonic event. The Santos action is merely the tremor.
If the rulemaking classifies political event contracts as contrary to the public interest, prediction markets lose their highest-volume, highest-attention category. Political contracts drove an outsized share of 2024 volumes, particularly on Polymarket. Removing that category forces platforms into a search for alternative event classes—sports, economics, climate, entertainment—with uncertain economics. The platforms that diversify their event pipelines now will survive the shock. Those that remain politically concentrated will not.
The second vector is user behavior. Prediction markets are disproportionately popular with U.S. users. Every manipulation case involving a U.S. politician accelerates the perception that the sector is a regulatory minefield. Traditional financial and sports-betting heavyweights monitoring the space will factor this into their entry decisions. A few more enforcement actions like this one, and the institutional timeline stretches further.
The winners in this cycle will be the platforms that build compliance into their architecture before the hammer falls. The losers will be the ones that treat $35,000 as noise, when it is actually a signal written into the regulatory ledger. The next manipulation case will be bigger, and the penalty will not be symbolic. The only question is whether the market structure evolves before that case finds its defendant.