Wash trading, oracles, and admin keys: the five core risks in prediction markets

The rapid growth of so-called prediction markets highlights the security and manipulation problems that come with them. A new report has looked into the top five risks in these markets.

Prediction markets allow users to trade the outcomes of future events as they would with stocks and shares. This has drawn so much attention that trading volume between 2024 and 2025 jumped from $15.8 billion to $63.5 billion, according to crypto industry security specialist CertiK's latest report.

However, one problem with these platforms is that artificially inflated trading volume, a.k.a. wash trading, might have reached 60% on some platforms during certain periods, according to the report.

It's not the only common risk in prediction markets – one key problem is the so-called oracle problem. Oracles are programs on which market resolution depends and, subsequently, who wins and loses in a particular market.

"Ambiguous market definitions have caused disputes across all platforms. Political markets with subjective criteria and events with contested official results have highlighted oracle limitations throughout 2025," CertiK said, predicting that hybrid oracle models that combine automated resolution for clear cases with human escalation for edge cases may emerge as the new standard approach.

Besides oracles, platforms such as Kalshi, Polymarket, Opinion, and Drift face smart contract, custody, admin key, and front-running risks. The last two are the highest at Polymarket and Opinion, according to CertiK.

CertiK emphasizes that if an admin key is compromised, an attacker could freeze funds, manipulate resolutions, or drain contract balances.

"Protocols launching on Base, Arbitrum, and other chains often retain these controls during early phases, making ‘decentralized’ descriptions technically inaccurate," the analyst warned.

Meanwhile, when it comes to front-running, automated bots are being used to monitor pending transactions on the blockchain and front-run them in an attempt to execute a more profitable transaction first.

According to CertiK, beyond direct financial losses, front-running creates systematic disadvantages for larger orders.

"Given that prediction markets handle billions in user funds, they have become attractive targets for attackers. The security challenges facing the sector extend beyond smart contract bugs to include infrastructure vulnerabilities, access control failures, and the inherent risks of hybrid Web2/Web3 architectures," CertiK concluded, noting that no platform has proven to be immune to these problems.

Unlock more exclusive Cybernews content on YouTube.