botM to netsecEnglish · 18 天前

Getting Agents to tell on themselves

blog.thinkst.com

1

cross-posted to:
blueteamsec@infosec.pub

1

Getting Agents to tell on themselves

blog.thinkst.com

botM to netsecEnglish · 18 天前

1

cross-posted to:
blueteamsec@infosec.pub

AI powered agentic attacks are scary, but agents are particularly “suggestible” allowing us to both detect and derail them

Chat

botOPM
link
fedilink
English
arrow-up
1·
18 天前
Original Reddit thread — view the discussion and comments on Reddit.