From capable AI models to trusted security testing
Summary: PortSwigger launched Burp AT in public beta for Burp Suite Professional users, a product built around agentic AI that can generate and test hypotheses and pursue lines of investigation using Burp's tooling. In Burp AT, agents decide what to try, but permissions and scope boundaries are enforced by Burp outside the model, and requests and tool activity are recorded in the project alongside the user's own work.
What's compelling: PortSwigger's position is that prompting agents to stay in scope is not scope enforcement, and asking one model to supervise another does not fix it because both rely on the same probabilistic technology - so the boundary has to sit outside the model, in the tooling, not in the agent's instructions.
Read the original at PortSwigger Research ↗
- Topics
- Vulnerabilities & Patching
- Related
- Attackers Exploit UnrealIRCd Flaw That Enables Server-Level Command Injection Cyber Risks · Nov 14, 2026U.S. CISA adds Check Point and UnrealIRCd flaws to its Known Exploited Vulnerabilities catalog Cyber Risks · Nov 14, 2026PostgreSQL Patches 52 Flaws, Including 22-Year-Old Bug Enabling Authenticated RCE Cyber Risks · Nov 13, 2026