AI agents aren't just chatbots anymore. They're getting tools, permissions, and access to sensitive data. 😳 In this @yow_conf talk, @InsiderPhD explores how malicious prompts can hijack agents—and turn them into an insider threat. Watch: piped.video/fuJq99juh0Q?list=PL…

Aug 18, 2026 · 12:06 PM UTC

1
2
5
768
Sort replies: Relevant Recent Liked
Agent permissions turn malicious prompts into security incidents, not mere bad outputs.
7