Running work in parallel - Port Swigger

Summary: PortSwigger's Burp AT can investigate different parts of a target simultaneously, splitting work either through delegated AI agent workers that report results back into the task's conversation or through sub-tasks that run in separate conversations. In Manual or Smart mode Burp AT asks for approval before using workers, while in Autonomous mode it can use them without asking, and approving workers does not approve their tool calls.
Bottom line: Workers can keep running after you clear or replace a goal, and the only remedy the documentation offers is clicking Stop - so an autonomous scanning agent can still be acting on your target after you believe you have redirected it.

Read the original at PortSwigger Research ↗

Topics: Vulnerabilities & Patching
Related: Attackers Exploit UnrealIRCd Flaw That Enables Server-Level Command Injection Cyber Risks · Nov 14, 2026
U.S. CISA adds Check Point and UnrealIRCd flaws to its Known Exploited Vulnerabilities catalog Cyber Risks · Nov 14, 2026
PostgreSQL Patches 52 Flaws, Including 22-Year-Old Bug Enabling Authenticated RCE Cyber Risks · Nov 13, 2026