PortSwigger Launches Burp AT: Agentic AI for Penetration Testing

PortSwigger has launched Burp AT in public beta, integrating agentic AI into Burp Suite Professional. This feature allows penetration testers to delegate specific tasks to AI agents while maintaining control over scope and conclusions. The AI can analyze complex code, such as minified JavaScript, and identify vulnerabilities that may go unnoticed otherwise. In a closed beta test, Burp AT helped uncover a critical vulnerability in a substantial codebase that would have remained untested for a year. The tool is designed to enhance human-led pentesting workflows, allowing professionals to focus on high-priority tasks. Burp AT is available now for Burp Suite Professional users, with plans for future enhancements and more autonomous testing capabilities.

Key Points: • Burp AT integrates agentic AI into Burp Suite Professional for enhanced pentesting. • Penetration testers can delegate tasks to AI while retaining control over the process. • A closed beta revealed a critical vulnerability in 66,000 lines of code that would have been missed.