Autonomous AI agents are no longer theoretical cyber risks as new research shows a large language model executing an end-to-end ransomware campaign, highlighting the urgency of patching known ...