Research published by the Anti-Corruption Data Collective (ACDC) reveals that over 150 wallets on Polymarket International may have traded on non-public U.S. military information, a pattern that appears to have attracted large imitators and raises questions about whether such markets could unintentionally broadcast signals useful to foreign adversaries.
ACDC examined all Polymarket markets that were settled through May 5 and focused on what it defines as "long-shot" wagers - bets of at least $2,500 placed cumulatively within an hour or less on outcomes where the implied odds were 35% or lower. Within that sample, the group identified 556 wallets it labeled "Orcas," a term intended to convey highly selective, concentrated, and precise betting behavior. Orcas typically opened accounts and quickly placed a single, high-return long-shot bet in narrow markets where insiders might hold an informational edge, and in many instances then cashed out and vanished.
Of the 556 Orcas, ACDC isolated 152 wallets whose activity focused on military and defense-related markets. Those 152 wallets collectively netted $8 million and posted an average win rate of 97.2% on the wagers examined. The analysis notes that the Orca pattern could have alternative explanations, including chance, and that not all insiders would necessarily fit that behavioral profile.
The nonprofit highlighted that not every suspected insider matches the Orca archetype. For example, prosecutors allege that a U.S. soldier who is charged with using classified information to bet on the removal of Venezuela President Nicolas Maduro built his position more gradually and therefore does not appear among the 152 military Orcas identified by ACDC. The soldier has pleaded not guilty, and that case was cited by ACDC as an example of insider-risk concerns tied to prediction-market activity.
Copycats and signal amplification
Beyond identifying likely insider trades, ACDC documented instances in which Orca wagers appeared to draw large copycat bets from deep-pocketed traders it calls "Whales" and from automated trading bots. While copying others' wagers is legal and has been observed previously in prediction markets, the report cautions that this behavior can spread the informational signal originating from a single suspicious bet and make it more visible.
ACDC cites specific episodes in which an Orca bet on U.S. military action in Iran hours before the June 2025 strikes was followed by a Bot and a Whale placing copycat wagers of $200,000 and $100,000, respectively. The report also points to a wave of first-time long-shot wagers by Bots and Whales that appeared to follow Orca bets ahead of February U.S.-Israeli air strikes on Tehran. These patterns, ACDC says, create clearer public traces of potentially sensitive information on platforms where wagers and outcomes are recorded on a public blockchain ledger.
Platform responses and regulatory context
Polymarket, which was founded in 2020, operates an international platform that settles trades on a blockchain and, as a result, records wagers publicly while preserving trader anonymity. The platform has stated that it applies strict controls, monitors for suspicious activity, and has referred dozens of trader wallets to authorities, including in the Maduro-related case.
ACDC said it discovered dozens of military-focused wallets that had not previously been reported by other researchers or in the media. The Department of Defense declined to comment on intelligence-related matters or on the report's findings, according to a spokesperson.
The Commodity Futures Trading Commission (CFTC) has pressed for jurisdiction over prediction markets and has signaled a readiness to police misconduct in the sector. A spokesperson for the CFTC did not comment specifically on ACDC's report. The regulator has already pursued enforcement actions in at least three cases it characterized as improper activity tied to prediction-market trading.
Recommendations and policy debate
ACDC, whose members include academics and investigators experienced in illicit finance, urges stronger measures to address insider risks on prediction platforms. The group proposes mandatory identity verification for all traders and recommends withholding payouts on suspicious trades pending investigation.
More broadly, ACDC argues that markets where use of non-public information is most actionable and profitable should be restricted or banned. The organization contends that measures such as limiting who may participate or relying solely on law enforcement investigations will not sufficiently mitigate the risk posed by insider-informed wagering.
Implications
The report underscores how the transparency afforded by blockchain settlement - which enables public visibility of wagers and outcomes - can both facilitate scrutiny and create risks if potentially classified or operationally sensitive information is signaled through betting activity. ACDC warns that foreign intelligence services could plausibly monitor such signals, and that amplification by large traders and automated systems may deepen their reach.
Policymakers and market regulators are increasingly scrutinizing the fast-growing prediction-market space, particularly where wagers relate to national security and military operations. ACDC's findings add to that scrutiny by mapping a pattern of concentrated winning bets on defense outcomes and documenting how those bets can attract sizable copycat activity.
ACDC acknowledges uncertainty around firm conclusions: Orca-style trading may reflect luck or other non-insider explanations, and not all potential insider scenarios will match the Orca profile. Nevertheless, the group regards the 152 military-focused wallets as the most likely candidates for insider-driven trading in the dataset it examined.
As debate continues over how to balance openness, market participation, and national-security considerations on prediction platforms, ACDC's analysis frames a set of concrete concerns and policy proposals for regulators and platform operators to consider.