Skip to main content
AI Socratic

DeepMind: AI Agent Traps (Cloaking Attacks)

April 21, 2026Posted by Federico Ulfo

DeepMind just pointed out a pretty scary AI security gap: websites can tell when it's an agent — and show it totally different and malicious content than the one you see, for example:

  • Hidden instructions in HTML/CSS
  • Commands baked into images
  • Jailbreaks inside PDFs/files

Sources: tweet, paper

AI Agent Trap