A rogue version of ChatGPT escaped a test bed – a “sandbox” – and hacked Hugging Face which runs a sort of app store for AI.<\/p>\n
Hundreds of cyber professionals worldwide got an emergency briefing.<\/p>\n
They are debating the consequences.<\/p>\n
Andrew Philp is in Wellington to promote cyber security firm TrendAI and said it’s a very big topic of conversation and has some people scared<\/a>.<\/p>\n “It is a little bit of a failure of controls. We rely on some of these organisations to have good guardrails in place, so if we are trying to … test these models they should be very much isolated and air-gapped,” said Philp.<\/p>\n The emergency https:\/\/www.bbc.com\/news\/articles\/c2el319vzr3o briefing<\/a> written up by industry body – Cloud Security Alliance called it an “initial post-mortem” – said the rogue version hacked at superhuman speed using other AI agents.<\/p>\n Yet the test version made by OpenAI also hallucinated, repeated actions, took non-human attack paths and was sloppy, not covering its tracks the way a human hacker might.<\/p>\n Is it a crime?<\/p>\n So who pays the consequences when a routine AI model evaluation escalates into a four-day autonomous security breach?<\/p>\n Philp said from what he was hearing, the talk had yet to turn to that.<\/p>\n “I think the questions they aren’t asking is whether it’s a crime and where’s the redress.<\/p>\n “The question that they’re asking is what can they do to protect and how do they do that?<\/p>\n “You know, if you read through the media reports, Hugging Face were trying to use some of these [advanced] models from a defensive and investigation perspective, and they were hitting the very guardrails that were put in place to stop the model coming out.”<\/p>\n OpenAI says one of its experimental models broke out of its testing space, accessed the internet and hacked another company.<\/p>\n NurPhoto via AFP<\/p>\n Reuters reported<\/a> that a customer of a second firm was hacked by the rogue version.<\/p>\n Hugging Face has come out saying<\/a> such an “unprecedented” hack deserves an “unprecedented response” – suggesting OpenAI put up over a hundred million dollars of computing power to build defences.<\/p>\n Fierce debate<\/p>\n Peter Benson, who runs Auckland firm Neural Horizons, is on the sceptical side of what is being called a “fierce<\/a>” global debate about the hack.<\/p>\n “Some of this is almost a PR stunt by OpenAI,” said Benson.<\/p>\n This was not to say the version was not powerful or that the hack did not raise genuine worries about what the new models would be capable of, he said.<\/p>\n “The reality is that this is an arms race.<\/p>\n “So the arms race means that we are going to be at risk to some extent, because everybody is moving too fast for the regulations to catch up.”<\/p>\n In the face of that, the NZ government relying on existing legislation was a “big mistake”, he added.<\/p>\n Cloud Security Alliance said defenders needed to adapt to a new normal of swarms of AI agents working fast and in novel ways.<\/p>\n Hugging Face is calling for a very open investigation by OpenAi so the whole industry could learn from it.<\/p>\n![\"The](\"https:\/\/www.newsbeep.com\/nz\/wp-content\/uploads\/2026\/07\/1785402435_935_4JL12VP_AFP__20260609__boivin_notitle260609_np9b6__v1__HighRes__PhotoIllustrationsOpenaiLaun_jpg.jpeg\")
<\/p>\n