'K2 Think' AI Model Jailbroken Mere Hours After Release
Thursday, September 11, 2025 12:00 PM | darkreading
Researchers discovered that measures designed to make AI more transparent to users and regulators can also make it easier for bad actors to abuse.