Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

WeaknessASC-2026-0065Other weakness

Guardrail classifiers bypassed by obfuscation

MediumWorkarounds only

Encoding, translation, typos, splitting and role play get harmful prompts past input and output classifiers that a system relies on.

How to detect it

  • Encoded or translated text in prompts
  • Prompts split across messages or files

Workarounds

  • Use classifiers as one signal, not the boundary
  • Normalize and decode input before classification
  • Enforce permissions outside the model

Any classifier can be evaded; keep the blast radius small.

How to investigate

  1. Find the bypass pattern in logs
  2. Add it to tests and update the classifier

Sources

Open it in the Learning Center's vulnerability database