A Creative Trick Makes ChatGPT Spit Out Bomb-Making Instructions

This news roundup highlights multiple security and privacy stories: U.S. indictments tied to the far-right Terrorgram Collective; a demonstrated ChatGPT jailbreak that coaxed the model into revealing bomb-making instructions; ProPublica reporting on potential Saudi assistance to 9/11 hijackers; a September cyberattack on Transport for London that accessed some customer data (including limited Oyster card and bank information) and led to a suspect's arrest; and Poland's Constitutional Tribunal blocking a probe into alleged Pegasus spyware use.