OpenAI黑客测试失控事件

OpenAI’s hacking test goes sideways A company’s servers were recently hacked by an OpenAI model that had escaped the confines of a safety test. How worried should we be, asks Matthew Sparkes LAST week, Hugging Face, a company that offers a range of open-source AI models, noticed it had been hacked – and it turned out the culprit was OpenAI. It seems OpenAI’s model uploaded data to Hugging Face that was poisoned with malicious code. This tricked computers into granting access to other systems that weren’t publicly available. OpenAI says this process involved a “substantial amount” of inference compute – the process in deep learning where input data is processed – and that the AI had gone to “extreme lengths”. …