2026 - April 16th

It is no secret that Artificial Intelligence can create hallucinations. i.e. Robots will bullshit you from time to time.

What if the bullshit was by design?

Although I’m work-shopping the name and proof of concepts, the theory behind creating misleading information to mitigate attacks, I believe, is worth exploring.

This has shades of deception engineering written all over it.

What if:

• An attacker receives a fake API key?
• An LLM feeds an attacker a fake internal system prompt?
• An attacker thinks they dumped data tables that contained nothing but fabricated information?

The process of intentionally misleading those attacking AI systems can place a tremendous tax on threat actors.