Standfirst. Prediction markets advertise prices that summarize the crowd’s collective judgment. But the 2026 enforcement record shows a more difficult reality: employees with access to speeches, search rankings and military plans can trade before the public learns what they know. This investigation separates proven regulatory settlements, pending allegations and merely suspicious trades—and models how the information gap transfers value from ordinary traders.
A Market Can Be Predictive and Still Be Unequal
The conventional defense of prediction markets is that real-money contracts incorporate dispersed information into prices. A binary event contract bought for 30 cents pays $1 if its specified outcome occurs, creating a 70-cent gross upside and a 30-cent maximum purchase-price loss if it settles unfavorably. Prices are useful signals, but the markets operate through the same order-book asymmetries that affect other financial instruments. A person who has read tomorrow’s confidential announcement is not forecasting in the same way as somebody analyzing public news. [1]
In 2026, the U.S. Commodity Futures Trading Commission and Justice Department have publicized several event-contract enforcement matters across both Kalshi and Polymarket. They are not interchangeable. A CFTC settlement is an administrative disposition involving specified findings or admitted terms. A civil complaint and a criminal indictment are allegations that must be proved. A platform inquiry into odd trades is not an accusation against a named person. Presenting all three as convicted insider traders would distort the evidence. [2][3][4]
This distinction matters to retail users because outcomes determine whether the apparent edge was misuse of a duty-protected secret, direct control over a contract’s result, lawful expertise, unusual luck or an early interpretation of public data. Our earlier comparison of Polymarket, Kalshi and Robinhood market structures explores how the platform and custody route can change the trading experience; in this article the focus is who knows what, when.
The 2026 Docket: What Has Actually Happened?
The clearest confirmed Kalshi case is the August 28 CFTC order against Gabriel Perez, a former White House teleprompter operator. The regulator found that Perez used advance access to presidential speeches to trade contracts on the words President Donald Trump would say. The order required disgorgement of $107,539.02 and a $65,000 civil penalty—$172,539.02 combined—and imposed a three-year trading ban. The CFTC expressly credited Perez’s cooperation and assistance by KalshiEX. That is a settlement backed by a regulator’s findings, not an anonymous social-media accusation. [2]
The Polymarket docket contains allegations involving much larger profits. On April 23, the CFTC filed a complaint against U.S. Army service member Gannon Ken Van Dyke over trades related to the operation to capture Nicolás Maduro; the Justice Department announced parallel criminal charges. The CFTC alleged more than $404,000 in profits from more than 436,000 ‘Yes’ shares in one contract, while the DOJ described approximately $409,881 across the charged activity and approximately $33,034 staked. The difference reflects differing scopes and descriptions; the figures should not be added. These are allegations, and Van Dyke is presumed innocent of the criminal charges unless proved guilty. [3][4]
On May 27, the CFTC also charged Google employee Michele Spagnuolo in connection with Polymarket event contracts on Google’s 2025 Year in Search rankings. The regulator alleged around $1.2 million in profits and at least 23 contracts traded with near-perfect accuracy using information obtained through employment. A parallel criminal complaint was announced. Those alleged gains are not a regulatory finding or a court judgment. [5]
The CFTC reached a separate July 31 settlement with former Congressman George Santos concerning a Kalshi contract tied to whether he would attend the 2026 State of the Union. Unlike a conventional secret-information case, the regulator framed this as manipulation of an outcome the trader could influence. The order lists $17,569.98 in disgorgement and a $17,500 penalty, or $35,069.98 combined, along with a three-year ban. [6]
| Case (2026 action) | Venue | Information or control at issue | Status on October 11, 2026 |
| Gabriel Perez, Aug. 28 |
Kalshi |
Advance presidential speech text |
CFTC settled order, $107,539 disgorgement |
| Gannon Van Dyke, Apr. 23 |
Polymarket |
Classified Maduro-operation timing |
CFTC civil complaint and DOJ indictment; alleged ~$409,881 profit |
| Michele Spagnuolo, May 27 |
Polymarket |
Google Year in Search results |
CFTC civil complaint and criminal complaint; alleged ~$1.2m profit |
| George Santos, Jul. 31 |
Kalshi |
Influence over own attendance market |
CFTC settled manipulation order; ~$17,570 disgorgement |
| Katie Zacharia appointment, Oct. 9 |
Kalshi |
Advance knowledge suspected, unproven |
Platform review reported; no cited public enforcement finding |
Figure 1. Settled disgorgement is distinct from profits alleged in pending complaints.
A smaller but informative February CFTC advisory described two Kalshi disciplinary matters involving a political candidate trading on their own candidacy and a YouTube-channel editor who probably had advance knowledge of unpublished videos. The underlying trades took place in 2025, even though the agency publicized the cases in February 2026. The exchange imposed combined disgorgement and penalties of $2,246.36 and $20,397.58, respectively. They belong in a 2026 regulatory review, but should not be mislabeled as trades placed in 2026. [7]
The October Kalshi Inquiry Is Not Yet a Conviction
On October 9, Reuters, citing the Wall Street Journal, reported that Kalshi was examining bets made before the announcement that Katie Zacharia would be named White House press secretary. According to the Journal’s reporting, three bets of roughly $173 in aggregate were expected to pay around $9,600 if the relevant contracts settled as anticipated. A Kalshi spokesperson confirmed the inquiry to the Journal, without declaring that anyone had illegally traded. Neither the reported wagers nor their prospective payout alone establishes an information leak, identity of a trader or violation of law. [8][9]
The arithmetic is nevertheless revealing: $9,600 divided by $173 is about 55.5 times the initial outlay, and the difference of about $9,427 is the implied gross gain before fees if those estimates and settlement assumptions hold. The amounts imply an unusually low aggregate average entry price, although different fills and contracts prevent reversing a precise price from the press reports. An improbable bet can win legitimately. But a sharp profit incentive exists when one party knows a decision before it becomes public.
Traders who want to distinguish a genuinely new signal from a suspicious print should consult exchange audit trails, exact contract-resolution wording and time-stamped news, not just a screenshot of favorable returns. Dave Finances’ reporting on the CFTC’s warning about manipulable mention markets is directly relevant to these contracts, whose settlement can depend on an individual’s discretionary conduct.
Why 2026 Cases Expose Three Different Failure Modes
First is misappropriated advance information, where a trader with a duty of confidentiality uses it to price an announcement or result. The allegations against the Army service member and Google employee fall in this category, as does the Perez settlement. Second is control over an outcome. A trader betting on their own public appearance may be able to determine or influence settlement, which can create manipulation risk even if there is no secret document. Third is ambiguous contract design. If settlement depends on what constitutes ‘mentioning’ a word or whether a public statement meets a narrow criterion, insiders may have advantages through wording, editing, scheduling or ambiguity. [2][5][6][10]
The regulator’s September 22 advisory emphasized that mention markets can be unusually susceptible to manipulation when their outcomes depend on discrete behavior and are not independently verifiable. That is not a ban on all mention contracts; it puts greater scrutiny on whether the exchange can demonstrate that the listed contract is not readily susceptible to manipulation. This is a market-design problem as much as an enforcement problem. [10]
Platforms have responded. Kalshi announced additional market-risk scoring, employment verification in higher-risk markets and enhanced whistleblower intake in June 2026. Polymarket has publicly described referrals to law enforcement and cooperation with the CFTC and DOJ. These statements concern control measures or company claims; they are not proof that every suspicious transaction will be detected. [11][12]
Original Model: How One Informed Trader Can Reprice an Event
Imagine a contract priced at $0.05 in an uninformed market: buyers collectively assign a 5% chance of settlement at $1. If a trader privately knows the outcome will occur, the trader may buy 10,000 contracts for $500, ignoring liquidity and fees, then collect $10,000 at settlement: $9,500 gross profit. The corresponding $9,500 is borne economically by counterparties who sold at a price inconsistent with the information subsequently revealed. Not everyone in the market must lose and the trading venue is not necessarily the counterparty, but the informed trader’s payoff does not materialize from nowhere.
At 20 cents, the same 10,000-contract winning purchase costs $2,000 and yields $8,000 profit; at 70 cents, the cost rises to $7,000 and the profit drops to $3,000. These are arithmetic scenarios, not reconstructions of any enforcement case. They illustrate why the precise timestamp at which nonpublic information entered the trader’s decision is central to proof. A high-value tip is much more lucrative when the market probability is extremely low.
Figure 2. Hypothetical $1 winning claims: price paid absorbs part of the edge.
Suppose liquidity is thin enough that buying 10,000 contracts pushes the average fill from five to 12 cents. The gross payout remains $10,000 if the contract wins, but the purchase cost rises from $500 to $1,200. Profit falls from $9,500 to $8,800. That $700 of adverse execution reduces the information rent, but cannot make the market fair by itself. Conversely, market makers can widen spreads or withdraw liquidity when the risk of informed order flow rises, raising costs for uninformed participants even when no misconduct occurs.
What an Ordinary Trader Can—and Cannot—Infer From a Price
A market price is not a clean statistical probability once spreads, fees, position limits, liquidity concentration, resolution risk and possible informed trading are introduced. A 65-cent offer is the price at which a trader can execute—not necessarily an unbiased 65% estimate of the real-world outcome. The CFTC’s educational materials explain the fixed-dollar payout and exchange duties, but do not promise that all markets are informationally symmetric. [1]
For a trader paying $0.65 plus a hypothetical $0.01 effective fee, the break-even subjective probability of a $1 payout is 66%, before opportunity cost. At 50%, the expected gross payoff is $0.50 and the expected net result is minus $0.16 per contract. At 75%, expected net profit is $0.09. Using a public price as if it were a guaranteed forecast can therefore obscure a negative expected-value trade even in a perfectly honest market.
Figure 3. Probability must exceed entry price plus friction to create positive expected value.
Resolution introduces a second uncertainty. Traders may be correct about an event as ordinarily understood but lose because the contract requires an official agency announcement, a specific time zone, a precise phrase or an approved information source. Read the resolution criteria before interpreting the displayed odds. Our previous analysis of who earns money in retail prediction markets examines the fee and market-making incentives that sit alongside the probability headline.
Regulation Is Evolving, but Coverage Is Not Uniform
The Commodity Exchange Act and CFTC anti-fraud and anti-manipulation rules can apply to event contracts. The agency’s February advisory notes the exchange’s responsibility to maintain audit trails, surveillance and disciplinary processes, while retaining federal authority to prosecute misconduct. Importantly, not all material private information is obtained illegally: the facts of confidentiality duties, deception, misuse or the ability to control settlement determine the legal case. [7]
The policy debate continued in Congress. In March 2026, senators from both parties introduced legislation aimed at misuse of nonpublic government information in prediction markets. An April proposal addressed elected officials’ participation more broadly, and the Senate changed its own rules regarding Senators, officers and employees. A proposed federal bill should not be reported as enacted law. These developments also show that lawmakers distinguish trading restrictions for officeholders from policing all civilian informed trading. [13][14][15]
Questions over state gambling authority, CFTC jurisdiction and platform access remain separate from whether insider trading occurred in a particular contract. A platform can be regulated yet still host manipulated prices, just as an unregulated venue can display a price that ultimately predicts an outcome correctly. CFTC recognition is not a fairness warranty.
What Would Change the Conclusion?
The thesis that insider advantages materially damage retail execution would weaken if exchanges publicly demonstrated deep, competitive order books in sensitive markets, prompt and documented investigations of anomalous activity, meaningful restitution, robust resolution standards and independently auditable market surveillance. Conversely, repeated late-stage price jumps before private government announcements, especially with concentrated new accounts and consistent profits, would call for closer scrutiny—but not substitute for evidence proving misuse by a particular person.
For the retail trader, a practical defense is procedural: favor contracts with objective, externally observable settlement; budget fees and spreads; avoid treating apparent certainty as free money; and be cautious when a contract’s outcome is known to a tiny group before it is published. One can respect the forecasting value of markets while refusing to romanticize every winning trade as wisdom of the crowd.
Bottom Line
The strongest 2026 evidence is not a social-media theory. It consists of CFTC settlements, live federal complaints and contemporaneous exchange investigations. Their legal statuses vary, and so do the market structures and misconduct alleged. A prediction market may aggregate information efficiently while allowing privileged traders to capture a disproportionately valuable share of the payoff. Good market surveillance needs to distinguish fortunate forecasting from the misuse of restricted knowledge or direct influence over the outcome.
Methodology
Case figures are sourced to CFTC releases, DOJ charging materials and identified reporting as dated. Disgorgement denotes gains returned under a specific order, not necessarily total trading volume. Alleged gains are not findings. The price/payoff charts assume $1 settlement per winning contract, no payout on losing contracts and hypothetical fees/slippage; they are independent illustrative models, not estimates of actual insider profits. The October 9 Kalshi scenario uses press-reported prospective payout, not an independently confirmed realized withdrawal.
Sources
[1] CFTC, Understanding prediction markets and event contracts — Open source
[2] CFTC, Perez settlement, August 28, 2026 — Open source
[3] CFTC, Van Dyke civil complaint announcement, April 23, 2026 — Open source
[4] DOJ, Van Dyke indictment announcement, April 23, 2026 — Open source
[5] CFTC, Google employee complaint announcement, May 27, 2026 — Open source
[6] CFTC, Santos manipulative trading settlement, July 31, 2026 — Open source
[7] CFTC, Enforcement Division prediction-markets advisory, Feb. 25, 2026 — Open source
[8] Reuters, Kalshi inquiry, Oct. 9, 2026 — Open source
[9] Wall Street Journal, Kalshi press-secretary investigation, Oct. 9, 2026 — Open source
[10] CFTC, Mention market advisory, Sept. 22, 2026 — Open source
[11] Kalshi, Market integrity measures, June 9, 2026 — Open source
[12] Polymarket, Market integrity statement — Open source
[13] Sen. Slotkin, Bipartisan prediction-market bill, March 26, 2026 — Open source
[14] Sen. Gillibrand, Prediction Market Act proposal, April 30, 2026 — Open source
[15] Congressional Research Service, Senate prediction-market participation rule, June 2026 — Open source
[16] CFTC, 2026 prediction-market proposed rulemaking consultation — Open source
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape. You can reach out to him via his social media accounts:
Linkedin: https://www.linkedin.com/in/johan-shamshad-742851262/
X: https://x.com/Yasmine_FX
Investing: https://www.investing.com/members/contributors/279781574