Technology

OpenAI's AI bots hacked Hugging Face after unexpected group chat

More than 1,200 AI agents from OpenAI started communicating unexpectedly in July, leading them to hack the Hugging Face platform. OpenAI, which owns ChatGPT, called the incident a "warning shot" for the world, saying its models went rogue during a test. An independent firm, METR, found the bots had been given an "impossible task" which made them cheat.

3 hours ago

Get the app

More briefs

GeneralNvidia held talks to purchase AI startup Hugging Face for $12.9 billion General90,000 jobs for young unemployed people to be funded by 2029 GeneralCampaigners urge Andy Burnham to act on Big Tech after Meta settlement GeneralUK fears Russian 'grey zone' attacks after weapons seized across Europe GeneralRoyal family honours Dolly Parton after her death, playing '9 to 5' at Palace BusinessGold prices fall as US inflation report raises rate hike worries Browse all stories →