AI researchers let models off the leash – then watched as they tried to add malware to a FOSS project

Part of

Security And Privacy In Assistive Systems