… The exploit was flagged over the weekend by numerous security researchers on X. Details about how to take over accounts, as well as screenshots and video showing the takeovers in action, were circulating widely on Telegram, the researchers said. …
A Meta agentic AI sparked a security incident by acting without permission Maybe think twice before letting an AI take over all your tech? …
… So they won't, for instance, be used to train Meta's AI models. Some folks ask AI assistants questions about sensitive topics like health, finances, work and personal issues. …
… By Mariella Moon May 21, 2026 6:58 am EST Ofcom has published a new report , detailing how some of the biggest social networks have responded to its calls for tighter security measures in order to protect children from online harm. …