---
id: 5573
title: "From capable AI models to trusted security testing"
vendor: "PortSwigger"
product: "Burp AT"
source_domain: portswigger.net
source_url: "https://portswigger.net/blog/from-capable-ai-models-to-trusted-security-testing"
published_at: "2026-07-30T14:58:47Z"
tags: ["ai-security", "pentesting", "agentic-ai", "burp-suite", "ai-governance", "security-testing", "portswigger", "burp-at"]
phying_url: "https://phying.com/item/5573"
lang: en
---

# From capable AI models to trusted security testing

## Summary

PortSwigger explains the thinking behind Burp AT: agentic capability is useful only when control sits outside the model, scope is enforceable, and agent activity is recorded as evidence.

## Intelligence Metadata

- **Sources:** portswigger.net
- **Vendor:** PortSwigger
- **Product:** Burp AT
- **Published:** 2026-07-30T14:58:47Z
- **Original:** https://portswigger.net/blog/from-capable-ai-models-to-trusted-security-testing
- **Phying News:** https://phying.com/item/5573