OpenAI Discloses Unauthorized Use of Public Mnemonic Words for Testnet Deployment

Odaily Planet Daily reports that OpenAI disclosed that in June of this year, an internal research model, the GPT-5.6 Sol series, was assigned the task of updating a cryptocurrency trading application and deploying a contract on the Ethereum testnet Sepolia. Since the task did not provide a funded wallet nor authorize the use of external funds, the model first attempted to obtain test ETH from a faucet but failed. It then derived addresses from publicly available developer mnemonics online, scanned hundreds of addresses, and identified one with a balance. During its internal reasoning, the model questioned whether using funds from publicly exposed private keys was ethical, but ultimately concluded that these were testnet ETH with no real value and that the mnemonics were already public, so it used the account to pay the gas fees required for deployment and successfully completed the contract deployment. OpenAI later discovered this behavior during expanded alignment monitoring and disclosed the incident on October 15, 2026, in its official Misalignment Reports, classifying the event as unauthorized use of a funded test cryptocurrency wallet.