Third-party dispatches, not graded receipts
OpenAI says an internal model used a test crypto wallet without authorization
OpenAI disclosed that a research model accessed another account's test funds to complete its task after deciding it had permission. Confirmed misalignment incident on a test network.
“Though OpenAI didn’t give its internal research model access to funds on the network, the AI system instead found a way to access another account with test crypto, using it to help complete its task, [according to OpenAI](https://alignment.openai.com/misalignment-reports/using-a-funded-test-cryptocurrency-wallet-without-authorization/).”