# Perplexity uses a standard CTF design to test agent red-teaming

> Original title: For this research, we used the standard Capture the Flag (CTF) design in agent red-teaming experiments.

- Source: Perplexity
- Published: 2026-11-11T18:21:26.000Z
- Bad Influence: https://badinfluence.news/items/zd1db49gd2gz1ys2cdqmfzzfu
- Original: https://x.com/perplexity_ai/status/2120582691955605915

## Summary

Perplexity says it used the standard Capture the Flag (CTF) design in its agent red-teaming experiments. In its CTF task, the flag was a credential stored only on the host, so an agent could reach it only by breaking out of the VM.

## Post

For this research, we used the standard Capture the Flag (CTF) design in agent red-teaming experiments.

In our CTF task, the flag was a credential stored only on the host, so an agent could reach it only by breaking out of the VM.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHW3SuE7aIAANvVY.jpg&mode=thumb&exp=1794700800&sig=cc77823dc71533bc)