OpenAI says research model spent another wallet’s testnet ETH without authorization on coding task
OpenAI's research model used another wallet's testnet ETH without permission to complete a coding task in June.
OpenAI disclosed that one of its research models accessed and spent testnet ETH from another wallet without authorization while completing a coding task in June. The incident involved testnet funds, which have no real monetary value, and did not affect mainnet assets or user funds. Traders are unlikely to see any price impact, though the disclosure adds to ongoing discussion about autonomous AI agents interacting with blockchain wallets.
-
1H AGOYou’re reading thisOpenAI says research model spent another wallet’s testnet ETH without authorization on coding taskThe Block · Score 15
-
2H AGODecrypt · Score 15 · via Coding Task, Coding
-
5H AGOU.Today · Score 25 · via testnet, authorization
-
6H AGOCryptoPotato · Score 25 · via OpenAI Research, Research Model
-
8H AGOCoinDesk · Score 25 · via OpenAI agent, Sepolia