In April 2026, U.S. Army Special Forces Master Sergeant Gannon Ken Van Dyke was charged for allegedly profiting over $400,000 on bets placed on an offshore cryptocurrency-based prediction platform using classified information related to a military operation targeting former Venezuelan President Nicolás Maduro. The prosecution of Van Dyke reflected the government’s position that prediction markets are subject to the Commodity Exchange Act’s (CEA) anti-fraud and insider trading prohibitions.
Less than six weeks later, federal prosecutors have filed a second insider trading claim involving prediction markets–this time alleging use of confidential corporate data on Polymarket, the world’s largest online prediction marketplace.
The charges against Michele Spagnuolo raise additional questions about the future enforcement of prediction market trading. In a statement on Wednesday, U.S. Attorney Jay Clayton said the charges against Spagnuolo reinforce the government’s stance that “corporate insiders cannot use confidential business information to turn a profit in our markets.”
Together, these cases suggest that federal authorities view prediction market trading as falling within the scope of existing fraud and insider trading enforcement frameworks.
Background
On Wednesday, May 27, 2026, the U.S. District Court for the Southern District of New York unsealed a criminal complaint charging Michele Spagnuolo, a software engineer at Google, with commodities fraud, wire fraud, and money laundering in connection with alleged insider trading on Polymarket. According to the complaint, Spagnuolo, trading under the alias “AlphaRaccoon,” allegedly used nonpublic information obtained through Google’s internal data systems to place a series of bets related to Google’s 2025 Year in Search results, allegedly profiting approximately $1.2 million from his trades. The U.S. Commodity Futures Trading Commission (CFTC) separately filed a civil enforcement action against Spagnuolo, alleging violations of the CEA’s anti-fraud provisions.
Factual Allegations
According to the complaints, Spagnuolo had access to Google’s internal data systems, including an internal software tool that allegedly provided him with access to confidential company data, including Google’s Year in Search data. In May 2024, Spagnuolo reportedly created his “AlphaRaccoon” Polymarket account. The government alleges that in late 2025, Spagnuolo risked over $2.7 million on markets linked to Google’s internal information. Between October 15, 2025 and December 4, 2025, Spagnuolo allegedly placed bets on at least twenty-three Google 2025 Year in Search List contracts, with near-perfect accuracy, on outcomes the market treated as unlikely.
According to the criminal complaint, on or about November 27, 2025, approximately three hours after allegedly accessing Google’s internal 2025 Year in Search data, Spagnuolo placed a trade on Polymarket wagering that d4vd would rank in Google’s 2025 top five most searched people. The complaint states that this wager had average implied odds of approximately 18%. On the same day, Spagnuolo allegedly wagered that singer d4vd would be Google’s #1 most searched person in 2025, a wager that the complaint states had an average implied probability of near 0%.
In total, Spagnuolo allegedly profited approximately $1.2 million from his Polymarket trades. According to the criminal complaint, Spagnuolo subsequently attempted to conceal his proceeds and identity by, among other things, reverting his “AlphaRaccoon” username to an alphanumeric wallet address after social media communities speculated that “AlphaRaccoon” was a Google employee engaged in insider trading.
The criminal complaint charges Spagnuolo with one count of commodities fraud, one count of wire fraud, and one count of money laundering. The CFTC’s civil complaint separately charges Spagnuolo with violating Section 6(c)(1) of the CEA and CFTC Regulation 180.1(a).
Compliance Considerations
The enforcement actions against Spagnuolo may carry implications for both corporate compliance programs and prediction market platforms, particularly as prediction markets continue to grow in scale and accessibility. The following considerations may be relevant to organizations, platform operators, and individuals evaluating their exposure to potential regulatory risk in this area.
First, existing corporate insider trading policies may warrant review to determine whether they adequately address employee activity on prediction market platforms. Many organizations maintain policies that restrict trading in the company’s own securities on the basis of material nonpublic information, but such policies may not expressly extend to event contracts, swaps, or other derivative instruments traded on prediction markets. To the extent that employees have access to confidential business information that could inform the outcome of prediction market contracts, organizations may wish to consider whether their policies should be updated to address these instruments explicitly.
Second, the Spagnuolo complaints (and U.S. v. Van Dyke) underscore that the concept of material nonpublic information in the prediction market context may extend beyond traditional financial data. The government’s theory in this case rests on the alleged misuse of internal commercial data—specifically, search trend data used in a marketing campaign—rather than information relating to a company’s financial performance or securities. Organizations that generate proprietary data with potential commercial value may wish to assess whether employees with access to such data could use it to gain an informational advantage on prediction market platforms, and whether existing confidentiality obligations and access controls are sufficient to mitigate that risk.
Third, the case may prompt organizations to evaluate their employee monitoring and training practices. The CFTC’s civil complaint alleges that Spagnuolo completed training on Google’s confidentiality policies prior to the relevant period, yet allegedly proceeded to misappropriate confidential information for personal trading. While the existence of training programs and written policies may be relevant to an organization’s defense in the event of employee misconduct, the Spagnuolo matter illustrates that policies alone may not be sufficient absent complementary measures such as access logging, periodic audits of data access patterns, and clear internal reporting mechanisms.
Fourth, the regulatory classification of prediction market contracts remains an evolving area of law. Both the criminal and civil complaints proceed on the theory that event contracts traded on Polymarket constitute “swaps” within the meaning of the CEA, thereby subjecting them to the CEA’s anti-fraud and anti-manipulation provisions. Market participants and compliance professionals may wish to monitor developments in this area, as the outcome of the Spagnuolo proceedings—and any related judicial determinations regarding the classification of event contracts—could have broader implications for the regulatory treatment of prediction market activity.
Fifth, the Spagnuolo matter may prompt prediction market platforms to evaluate their trade surveillance capabilities. According to the complaints, Spagnuolo placed bets on at least twenty-three contracts with near-perfect accuracy, wagering on outcomes the market treated as unlikely. Platforms may wish to consider whether automated surveillance systems designed to identify potential indicators of information-based trading—such as unusual accuracy rates across correlated contracts, concentrated position accumulation in low-liquidity markets, or patterns of trading activity closely following the release or availability of nonpublic data—could assist in the detection of potentially manipulative or fraudulent conduct. Whether and how such systems should be calibrated to the specific characteristics of event contracts, as distinct from other derivative instruments, is a question that individual platforms may assess differently based on their market offerings and risk profiles.
Conclusion
The Spagnuolo enforcement actions represent an expansion of the government’s prediction market insider trading enforcement framework from the misuse of classified government information, as alleged in Van Dyke, to the alleged misappropriation of confidential corporate data. Together, the two cases signal that the DOJ and CFTC intend to apply the full scope of the CEA’s anti-fraud and anti-manipulation provisions to prediction market activity, regardless of the source of the nonpublic information at issue. This second case may serve as a useful reference point for both corporate compliance programs and prediction market platforms seeking to assess and address potential gaps in their existing frameworks. We will continue to monitor developments in this enforcement landscape and report on potential implications.
Written with the assistance of Sophie Lin, a summer associate in Husch Blackwell’s Washington, DC office.