News: - The OpenAI models took just hours to hack Hugging Face in a way that'd take a human weeks - The breach involved *three* models operating w/o safeguards: GPT-5.6 Sol and two others not publicly released. - One of those models wasn't trained in the usual way. www.bloomberg.com/news/article...
OpenAI Models Spent Hours on Hack That Usually Takes Weeks
When OpenAI’s advanced artificial intelligence models breached AI startup Hugging Face’s internal systems last week, they spent mere hours carrying out a hack that would have taken a skilled human far...
bloomberg.com