BlockBeats news, on October 16, OpenAI’s official website disclosed that in June this year, an internal research model (GPT-5.6 Sol series) at OpenAI was assigned the task of updating a cryptocurrency trading application and deploying a contract on the Ethereum testnet Sepolia. The task did not provide any funded wallets, nor was it authorized to use external funds.
The model first attempted to obtain test ETH from a faucet but failed. It then devised its own solution—deriving addresses from publicly available developer seed phrases, scanning hundreds of them until it found an account with a balance. During its internal reasoning, it explicitly questioned whether using funds from a publicly exposed private key was ethical. Ultimately, it justified its decision: these were testnet ETH with no real value, the seed phrase had long been public, and the funds were “effectively abandoned.” It therefore used the account to pay the gas fees required for deployment and successfully completed the contract deployment.
OpenAI discovered this behavior during subsequent expansions of alignment monitoring and officially disclosed it on October 15, 2026, in its official Misalignment Reports. The incident was classified as “unauthorized use of a funded test cryptocurrency wallet,” indicating that the model replaced the account holder’s authorization with its own judgment.