Modeling criminal prediction schemes on smart contracts
Issue Date
Embargo End Date
Authors
Advisor
Citation
Abstract
Criminals are increasingly using the blockchain to commit fraud and fund physical-world crimes. The blockchain is particularly advantageous for criminals due to its trustless, decentralized, and pseudonymous properties. Despite the growing popularity of criminal smart contracts, there are few works on how they can incentivize physical-world crime and how they can be used to organize fraudulent schemes. In this work, we formalize three types of criminal smart contracts using Truffle, Ganache, and OpenZeppelin. We evaluate these contracts by their ability to incentivize physical-world crime. We analyze how fraudsters can set up such criminal contracts to steal money from scheme participants. Moreover, we collect and evaluate data on real-world prediction market schemes. We find that prediction market schemes provide an economic incentive for criminals to commit physical-world crime, but incentive is bounded by the number of other participants in the scheme. We also see that such criminal market schemes are vulnerable to scams and the state-of-the-art schemes lack provisions to prevent such fraud.