AI Agents Are Getting Scary

This is scary. AI agents getting more autonomous and accessing the internet raises some serious safety questions.

https://www.nytimes.com/2026/09/28/technology/openai-astra-safety.html?smtyp=cur&smid=tw-nytimes