Research from the Anti-Corruption Data Collective (ACDC) says at least 152 anonymous wallets on Polymarket may have traded using advance knowledge tied to U.S. military and defense developments. The group said those wallets generated about $8 million in profit and posted an average win rate of 97.2%.
ACDC also said blockchain data on its own cannot prove who controlled the wallets or where any information came from. For that reason, the research does not by itself establish that insider trading took place.
The group reviewed settled Polymarket markets and screened for what it called low-probability, high-value bets. Its filter focused on trades in which at least $2,500 was wagered within one hour on outcomes priced at no more than 35%.
Using that method, ACDC identified 556 wallets with unusual trading patterns. Of those, 152 were tied to military and defense markets.
Wallets labeled “Orcas” showed a recurring pattern
The researchers referred to the wallets as “Orcas.” According to the report, these addresses tended to appear suddenly, place bets on low-probability events, achieve unusually high success rates, and then leave after taking profits.
ACDC said success rates for low-probability wagers in military and defense markets were notably higher than the broader level seen across Polymarket. In some cases, the unusual trades were quickly followed by large traders and automated trading bots.
One example in the report involved the period before the U.S. strike on Iranian nuclear facilities in June 2025. After an unusual wager appeared in a U.S. military action market, one automated trading bot placed a $200,000 bet and another large trader put down $100,000. The researchers said they found similar unusual bets and copy-trading behavior before the U.S. and Israel struck Tehran.
Public on-chain activity may already be under close watch
David Szakonyi, co-founder of ACDC, said suspicious trading activity on Polymarket is more visible than many traders realize. He said large traders and bots are already tracking and copying suspected informed trades, and added that foreign intelligence agencies could also be monitoring these public on-chain transactions.
Regulators have been watching prediction-market abuse
Questions around insider trading in prediction markets have drawn continued regulatory attention. In April, the U.S. Commodity Futures Trading Commission (CFTC) accused U.S. Army servicemember Gannon Ken Van Dyke of using confidential information about the capture of former Venezuelan President Maduro to trade Polymarket contracts. The agency said he made more than $404,000. That case is not directly linked to the 152 wallets identified in the new research.
In May, the CFTC also charged a Google software engineer with trading 23 Polymarket contracts using confidential information tied to the company’s 2025 “Year in Search” rankings. The agency said the trader made about $1.2 million with near-perfect accuracy.
Polymarket has previously said it closely monitors suspicious trading and has reported dozens of wallets to the relevant authorities. As prediction markets continue to expand, insider trading, market manipulation, and national security risks are becoming key reasons for tighter regulatory scrutiny.