An AI Just Learned to Hack on Its Own — and It's Not Being Released

Bottom Line

New AI model finds and exploits security flaws autonomously, raising stakes for every connected system.