When Everyone Is an Insider:

How Prediction Markets Broke Insider Trading Law

The legal framework that took eighty years to build around corporate insiders was never designed for a world where the underlying asset is reality itself, and the platforms enforcing it cannot tell the difference between a cheat and a card counter.

Insider trading law in the United States rests on a simple premise: that the people who possess material nonpublic information can be identified, and that their obligation not to trade on it can be traced to a definable legal duty. That premise held for eight decades because the information that mattered was corporate information, and corporations have identifiable insiders. Prediction markets have upended that premise. When the contract settles based on whether a war starts or a president is deposed, the "insider" is no longer a corporate officer or an investment banker. It could be anyone: a soldier, a staffer, a campaign volunteer, a spouse. The legal framework has no answer for this.

It also has no answer for what happens to the people who are simply good at analysis. Prediction market platforms, lacking the doctrinal tools to distinguish between traders who possess stolen secrets and traders who are just better at reading publicly available information, have defaulted to the one metric they can measure: who is winning. That is not a regulatory framework. It is a tell that no framework exists.

Insider Trading on Kalshi, Polymarket, and the CFTC's Response

In the span of a few weeks this spring, a U.S. Army Special Forces soldier was indicted for using classified intelligence about the military extraction of Venezuelan President Nicolás Maduro to earn over $400,000 on Polymarket. Kalshi suspended and fined three congressional candidates for betting on their own political races. A video editor for one of YouTube's most popular creators was disciplined for trading on advance knowledge of upcoming content. And the CFTC issued its first-ever enforcement advisory on insider trading in prediction markets, asserting "full authority to police illegal trading practices occurring on any DCM."

The press covered these as scandals. Congress treated them as emergencies. The platforms treated them as proof that self-regulation works. All of them missed the deeper problem. These cases were not evidence that prediction markets have an insider trading problem that can be policed within the existing framework. They were evidence that the existing framework does not fit.

Why Securities Insider Trading Law Does Not Fit Prediction Markets

Securities insider trading law, developed over eighty years through cases like Chiarella, Dirks, O'Hagan, and Salman, rests on a structural feature that is easy to take for granted: the universe of potential insiders is knowable and bounded. Material nonpublic information relates to a corporate issuer with defined relationships. The duty not to trade arises from identifiable sources: fiduciary obligations, employment agreements, contractual confidentiality. The information flows through traceable channels. When suspicious trading appears ahead of a corporate event, enforcement agencies can map the relationships and identify who had access. The rules are prospective. Participants can conform their conduct.

In 2010, Dodd-Frank gave the CFTC a comparable anti-fraud tool. The agency finalized Rule 180.1, modeled on SEC Rule 10b-5, and announced it would be "guided, but not controlled" by the body of 10b-5 judicial precedent. That is, it borrowed the SEC's most powerful enforcement weapon while reserving the right to diverge from the doctrinal guardrails that decades of litigation built around it. The early cases under Rule 180.1 fit the framework comfortably: an energy company employee trading his employer's commodities in a personal account. The duty was identifiable. The breach was clear.

Prediction markets present a fundamentally different problem. In securities, the underlying asset is the company. In prediction markets, the underlying asset is reality. When a trader buys a contract on whether Maduro will be removed from power or whether the Federal Reserve will cut rates, the information that resolves the contract is not generated by a corporate issuer with defined relationships. It is generated by governments, political campaigns, media companies, central banks, and military commands. Each has its own internal duty structure, but none of it maps cleanly onto the corporate-issuer model that insider trading law was built around.

Who Owes a Duty Not to Trade on a Prediction Market?

The misappropriation theory, which the CFTC enforcement chief identified at NYU Law School in March 2026 as the applicable framework for prediction markets, requires that the trader breach a duty of trust and confidence owed to the source of the information. Some of the recent cases satisfy that element without much difficulty. The Army soldier owed a duty under military classification rules. The YouTube editor likely owed a duty to his employer under an employment agreement. These are the easy cases.

The hard cases are where the analysis breaks down. Consider the campaign staffer who bets before internal polling drops. Does she have a pre-existing duty of trust and confidence? To whom? The campaign? The candidate? What if there is no NDA? What if she is a volunteer? A campaign staffer told NPR this spring that she and her colleagues "routinely placed bets on prediction markets before internal poll results went public." The CFTC would need to establish that trading on that information breached a specific duty owed to a specific source, not merely that the information was nonpublic and the trading was profitable.

Now consider the cases further along the spectrum. The congressional staffer who overhears a markup discussion. The Pentagon contractor's spouse who hears about a deployment over dinner. The retired intelligence officer whose thirty-year career gives her the ability to read publicly available signals with greater accuracy than other participants. At each step further from formal employment and contractual duty, the misappropriation theory's requirement of a pre-existing duty to the source becomes harder to identify, harder to prove, and harder to distinguish from a lawful analytical advantage.

Both the SEC and the CFTC have been explicit that insider trading law does not prohibit all trading on material nonpublic information. It does not create a parity-of-information regime. Derivatives markets have always allowed participants to trade on lawfully obtained information edges. The grain trader scouting crop conditions, the macro analyst reading satellite imagery, the geopolitical specialist monitoring shipping traffic through the Strait of Hormuz: these are legitimate advantages. The question is where the line falls between legitimate analytical edge and unlawful misappropriation. In securities, the corporate disclosure framework provides a relatively clean boundary. In prediction markets, no equivalent boundary exists. And neither the CFTC nor the courts have drawn one.

Win Rate as Guilt: The Card Counter Problem

Because prediction market platforms cannot easily distinguish between genuine insiders and superior analysts, the primary surveillance mechanism appears to be outcome-based. Flag the winners.

In sports betting, sportsbooks routinely limit or ban "sharp" bettors who win too consistently. The sportsbook has no obligation to explain its reasoning, no due process requirement, and no articulable standard for what win rate triggers exclusion. It is a private business managing its risk. Prediction market platforms, despite operating as CFTC-regulated Designated Contract Markets, appear to be importing this same model. Win rate triggers scrutiny. Sustained profitability creates a presumption of impropriety. The platform works backward from outcome to infer misconduct.

This methodology cannot distinguish between three fundamentally different categories of participant. The genuine insider, who has material nonpublic information obtained through a relationship that carries a duty. The superior analyst, who builds better models, processes public information faster, or has deeper domain expertise, and whose lawful information advantage the CFTC has explicitly said Rule 180.1 does not prohibit. And the lucky participant who hits a streak in binary-outcome contracts where even a modest sample of correct predictions can look anomalous. If the platform's primary detection metric is win rate, it collapses all three into one.

The practical consequences are serious. Market participants have no articulable rules to which they can conform their conduct before placing a trade. In securities, a corporate insider knows the rules: she files Form 4, she observes blackout periods, she does not trade ahead of material disclosures. On a prediction market, what is the rule? Do not win too much? Do not be right too often on geopolitical contracts? Do not concentrate your trading in markets where you happen to have expertise? No participant can conform her conduct to a standard that is defined only retroactively, after the platform has reviewed her P&L.

When a DCM suspends and fines a trader based primarily on statistical performance rather than relational evidence linking the trader to a source of MNPI, the trader is left proving a negative. Prove you did not have inside information. Prove you were just good at analysis. That is not a due process regime that any practicing lawyer would recognize as adequate.

And the chilling effect is where this does the most damage. If sophisticated participants learn that sustained profitability triggers investigation and potential suspension, the rational response is to limit trading, reduce position sizes, or leave the market entirely. The participants who exit first are the informed, analytical traders whose participation makes the market informationally efficient. What remains is noise.

If Winning Is Suspicious, the Market Cannot Function

Here is the deeper irony. Prediction market platforms defend their regulatory status, as CFTC-regulated derivatives rather than illegal gambling, on the Hayekian claim that they produce superior information aggregation. Decentralized knowledge, expressed through prices, generates forecasts more accurate than any individual analyst or institution. But the Hayek mechanism depends on participants with heterogeneous information and analytical ability being rewarded for accuracy. Penalize the most accurate participants and you degrade the informational efficiency that justifies the market's existence. The market devolves into a platform where everyone is equally uninformed. That is precisely the characterization that state attorneys general are advancing in their lawsuits against Kalshi and Polymarket.

Dodd-Frank Did Not Define Insider Trading for Prediction Markets

The question that exposes the entire structure is whether Dodd-Frank actually defines insider trading for commodity and prediction markets. It does not. Section 753 gave the CFTC a broad anti-fraud prohibition. Rule 180.1 implemented it in language modeled on 10b-5. But neither the statute nor the rule establishes elements, defines boundaries, or creates a framework that market participants can know in advance. Neither did 10b-5, of course. But the SEC had eighty years of adversarial litigation, appellate review, and Supreme Court correction to build the doctrine that today makes securities insider trading law knowable. Chiarella was decided in 1980, forty-six years after the Exchange Act. O'Hagan came in 1997. Salman in 2016. The guardrails were built case by case over decades.

The CFTC received the same broad tool sixteen years ago and has not built equivalent doctrine. There is no body of appellate case law defining what insider trading means in derivatives or prediction markets. The enforcement chief's March 2026 NYU remarks were a roadmap for how the agency intends to apply Rule 180.1, not a report on doctrine that has been tested and refined through judicial review. In the absence of developed doctrine, enforcement defaults to the metrics that are operationally available: win rate, profitability, timing. The CFTC's February 2026 enforcement advisory described two Kalshi fact patterns and restated the agency's authority. That is two anecdotes and an assertion of jurisdiction. It is not a regulatory framework.

The SEC's insider trading framework took eighty years to build. The CFTC is sixteen years in and has barely started. The prediction markets are not going to wait.

Sources

Chiarella v. United States, 445 U.S. 222 (1980).

Dirks v. SEC, 463 U.S. 646 (1983).

United States v. O'Hagan, 521 U.S. 642 (1997).

Salman v. United States, 580 U.S. 39 (2016).

Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 753 (2010).

CFTC Final Rule, "Prohibition on the Employment, or Attempted Employment, of Manipulative and Deceptive Devices," 76 Fed. Reg. 41,398 (July 14, 2011).

CFTC Enforcement Division Advisory on Prediction Markets Insider Trading (Feb. 25, 2026).

Remarks of CFTC Enforcement Director Brian Miller, NYU Law School, "CFTC Enforcement Priorities, Insider Trading in the Prediction Markets, and Cooperation with the CFTC" (March 31, 2026).

Congressional Research Service, "Prediction Markets and Insider Trading Law," LSB11406 (March 18, 2026).

United States v. Van Dyke (S.D.N.Y.) (DOJ indictment, Polymarket insider trading, classified intelligence re Venezuela).

KalshiEX LLC, Disciplinary Actions: Congressional Candidate Trading Suspensions (April 2026); YouTube Editor Insider Trading Suspension (February 2026).

NPR, Report on Campaign Staffer Prediction Market Trading (May 2026).

De Silva, R. Tamara, "When the Trade Comes Before the Tweet: The CFTC's Insider Trading Investigation," De Silva Law Offices (April 16, 2026).

De Silva Law Offices is a financial regulatory law firm in CFTC/NFA compliance, securities law, derivatives, and event contracts. She is a former futures floor trader and the author of a seven-article event contracts series including the predecessor to this article, "When the Trade Comes Before the Tweet: The CFTC's Insider Trading Investigation."