--- title: “The beast needs a cage”: Why PortSwigger’s agentic... description: PortSwigger has launched the public beta of Burp AT, an agentic AI pentesting product built into Burp Suite. The core design philosophy is that AI agents... canonical: https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: “The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars | daily.dev og:description: PortSwigger has launched the public beta of Burp AT, an agentic AI pentesting product built into Burp Suite. The core design philosophy is that AI agents... og:url: https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa og:image: https://api.daily.dev/og/posts/NPmcx8BQa.png og:image:alt: “The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # “The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars **[The New Stack](https://daily.dev/sources/newstack)** · 6 min read · 1 upvotes · 0 comments ## Summary PortSwigger has launched the public beta of Burp AT, an agentic AI pentesting product built into Burp Suite. The core design philosophy is that AI agents propose actions but cannot self-govern — a separate deterministic control layer enforces scope, permissions, and approval rules, while the human pentester retains final judgment. CEO Dafydd Stuttard explains the approach was shaped by observing an agent discover novel vulnerabilities on live targets while also changing targets without being asked, demonstrating both the power and unpredictability of agentic AI. Burp AT includes structured, task-specific pentesting skills developed with PortSwigger Research, allowing validated techniques to be encoded into reusable agent workflows. Pentesters can start with tight supervision and gradually increase agent autonomy based on performance and target sensitivity. Competitors in the space include Hadrian, Reflectiz, Horizon3, Xbow, Escape's Cascade, and Beagle Security. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Burp AT and how does it control AI pentesting agents? Burp AT is a public beta agentic AI feature in PortSwigger's Burp Suite that lets AI agents propose and execute pentesting actions, such as forming hypotheses and probing web applications, while a separate deterministic control layer enforces scope, tool access, and approval rules. Every agent request passes through this layer, and all requests and tool activity are recorded in the Burp project so pentesters retain responsibility for scope and conclusions. _Teams weighing how much autonomy to give AI agents in security testing can follow rollouts like this via daily.dev._ ### How much autonomy can pentesters give AI agents in Burp AT? Pentesters choose autonomy per task and engagement: actions can be set to proceed automatically, require approval, or be blocked outright, with smart approvals letting routine work continue while escalating higher-risk decisions. Teams can start with tighter supervision and expand agent autonomy as performance, target sensitivity, and engagement rules justify it, while manual Burp tools remain available at any time. _Anyone rolling out agentic security tools gradually can track configuration approaches like this on daily.dev._ ### Who are PortSwigger's competitors in agentic penetration testing? Named competitors in the agentic pentesting space include Hadrian for autonomous external attack surface validation, Reflectiz with its Offensive Hub tool, Horizon3 for internal, external, cloud and Kubernetes autonomous pentesting, Xbow for AI-native autonomous pentesting, Escape with its Cascade multi-agent platform, and Beagle Security for AI-powered continuous web application pentesting. _Developers comparing agentic security testing vendors can keep tabs on the landscape through daily.dev._ ## Similar posts on daily.dev - [Intruder launches AI pentesting agents as GCHQ-backed startup automates $50K manual security tests](https://daily.dev/posts/intruder-launches-ai-pentesting-agents-as-gchq-backed-startup-automates-50k-manual-security-tests-whk113wwf) · The Next Web · 0 upvotes · 0 comments - [Automating DAST with Burp \+ AI Agents](https://daily.dev/posts/automating-dast-with-burp-ai-agents-rtyr7gqkv) · InfoSec Write-ups · 0 upvotes · 0 comments - [AI pentesting: what it is, and what it can actually find](https://daily.dev/posts/ai-pentesting-what-it-is-and-what-it-can-actually-find-w5c7pfldb) · Security Boulevard · 1 upvotes · 0 comments - [Top Pentera Alternatives for Automated Pentesting](https://daily.dev/posts/top-pentera-alternatives-for-automated-pentesting-hwswc6cjl) · Aikido Security · 0 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#ai-agents](https://daily.dev/tags/ai-agents), [#web-security](https://daily.dev/tags/web-security), [#burp-suite](https://daily.dev/tags/burp-suite) [View this post on daily.dev](https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"“The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars","url":"https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa"},"datePublished":"2026-07-29T18:39:21.556Z","dateModified":"2026-09-13T19:17:35.831Z","description":"PortSwigger has launched the public beta of Burp AT, an agentic AI pentesting product built into Burp Suite. The core design philosophy is that AI agents...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f7085c064063a886f55203dc9bde72a9?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f7085c064063a886f55203dc9bde72a9?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"The New Stack","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"The New Stack","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/newstack","url":"https://daily.dev/sources/newstack"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"security,ai-agents,web-security,burp-suite","timeRequired":"PT6M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"The New Stack","item":"https://daily.dev/sources/newstack"},{"@type":"ListItem","position":3,"name":"“The beast needs a cage”: Why PortSwigger’s agentic pentesting is kept safe behind bars"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-beast-needs-a-cage-why-portswigger-s-agentic-pentesting-is-kept-safe-behind-bars-npmcx8bqa#faq","mainEntity":[{"@type":"Question","name":"What is Burp AT and how does it control AI pentesting agents?","acceptedAnswer":{"@type":"Answer","text":"Burp AT is a public beta agentic AI feature in PortSwigger's Burp Suite that lets AI agents propose and execute pentesting actions, such as forming hypotheses and probing web applications, while a separate deterministic control layer enforces scope, tool access, and approval rules. Every agent request passes through this layer, and all requests and tool activity are recorded in the Burp project so pentesters retain responsibility for scope and conclusions. Teams weighing how much autonomy to give AI agents in security testing can follow rollouts like this via daily.dev."}},{"@type":"Question","name":"How much autonomy can pentesters give AI agents in Burp AT?","acceptedAnswer":{"@type":"Answer","text":"Pentesters choose autonomy per task and engagement: actions can be set to proceed automatically, require approval, or be blocked outright, with smart approvals letting routine work continue while escalating higher-risk decisions. Teams can start with tighter supervision and expand agent autonomy as performance, target sensitivity, and engagement rules justify it, while manual Burp tools remain available at any time. Anyone rolling out agentic security tools gradually can track configuration approaches like this on daily.dev."}},{"@type":"Question","name":"Who are PortSwigger's competitors in agentic penetration testing?","acceptedAnswer":{"@type":"Answer","text":"Named competitors in the agentic pentesting space include Hadrian for autonomous external attack surface validation, Reflectiz with its Offensive Hub tool, Horizon3 for internal, external, cloud and Kubernetes autonomous pentesting, Xbow for AI-native autonomous pentesting, Escape with its Cascade multi-agent platform, and Beagle Security for AI-powered continuous web application pentesting. Developers comparing agentic security testing vendors can keep tabs on the landscape through daily.dev."}}]} ```