A new analysis by the nonprofit Anti-Corruption Data Collective (ACDC) has identified more than 150 cryptocurrency wallets on the prediction platform Polymarket that appear to have profited from non-public U.S. military information.

The research, which examined all settled markets through early May, found that 556 wallets exhibiting unusually precise trading patterns were placing high-stakes wagers in niche markets where insiders typically hold information advantages. Researchers termed these accounts “orcas”: they generally opened positions rapidly, withdrew funds shortly after successful bets, and often left no further trace.

Among this group, 152 wallets focused specifically on military and defense markets, collectively generating approximately $8 million in profits while maintaining an average success rate of 97.2%. These wallets typically placed substantial bets of at least $2,500 within a single hour on outcomes with implied probabilities of 35% or lower — a pattern suggesting access to information beyond public knowledge. While ACDC acknowledged that some results could theoretically be attributed to chance, the concentration of success in intelligence-sensitive markets points toward likely insider activity.

Polymarket, which operates on a public blockchain that records all transactions while preserving user anonymity, has stated that it maintains rigorous monitoring systems and has referred numerous suspicious accounts to law enforcement authorities. The platform has faced heightened scrutiny since April, when a U.S. soldier was charged with using classified information to wager on Venezuelan political outcomes.

That backdrop helps explain why the ACDC findings are drawing attention beyond Polymarket itself: prediction markets are designed to aggregate information quickly, but that same mechanism can also make unusual trading patterns easier to spot when the underlying topic is sensitive and the trader’s edge is not public. In that setting, the question is not only whether a given bet was profitable, but whether the market is being used in ways that intersect with national security concerns.

Copycat Activity and National Security Risks

The findings raise broader concerns beyond individual misconduct, as evidence suggests that successful military bets trigger significant copycat trading from automated bots and high-capital participants known as “whales.” When one identified account wagered on U.S. military action against Iran shortly before the June 2025 strikes, an automated trading bot and a whale subsequently placed copycat bets of $200,000 and $100,000, respectively.

Similar patterns emerged preceding February’s U.S.-Israeli air operations over Tehran, where first-time long-shot positions by bots and whales appeared immediately after the initial suspicious wagers.

This dynamic creates a mechanism through which potentially classified information becomes publicly visible through market activity, raising concerns that foreign intelligence services could monitor and exploit these signals to anticipate military operations. The Commodity Futures Trading Commission, which is actively seeking regulatory authority over prediction markets, has indicated it will strictly police misconduct in the sector and has already pursued charges in at least three cases.

ACDC has proposed several structural policy responses, including mandatory identity verification for all platform participants, temporary withholding of payouts on suspicious transactions pending investigation, and outright prohibition of markets where non-public government information could yield decisive trading advantages.

The organization argues that relying solely on post-hoc law enforcement investigations, or restricting categories of participants, will be insufficient to address the systemic vulnerabilities exposed by these trading patterns.

Source: Metaverse Post