Escape – The Offensive Security Blog

AI vs AI: How Cascade exploited an AI agent in production

We pointed Cascade, Escape's AI pentesting engine, at a production AI agent that a large organization runs internally. The agent had a guardrail built to catch prompt injection. Cascade got past it on the second try. The trick wasn't a smarter payload, just a different pretext,

Application Security

Authenticated scanning behind OAuth, MFA, and CAPTCHA

Authenticated vulnerability scanning reaches where your real risk lives, behind the login, and it is the kind of security scanning most tools fail at. A scanner hits the OAuth redirect, the MFA prompt, or the CAPTCHA, falls back to crawling as an anonymous visitor, and still finishes green, so the

Offensive Security

Continuous penetration testing: what it means when the testing never stops

In late 2025, Anthropic disclosed the first documented cyberattack run largely by AI, a state-linked group that used Claude Code to run 80 to 90 percent of an espionage campaign on its own, at thousands of requests a second. Offense now moves at machine speed while most defenses stay scheduled

Offensive Security

Penetration testing as a service (PTaaS), explained

Penetration testing as a service (PTaaS) made buying a pentest almost frictionless. You scope it in a portal, watch findings land on a live dashboard, and click a button to retest the fix, all without a single procurement call. But the test still runs on a schedule, so weeks after

Announcement

Escape joins Anthropic’s Cyber Verification Program to advance AI-powered offensive security

Good news doesn't arrive alone. This month we joined not one frontier AI program but two! Yesterday we shared that Escape joined OpenAI's Trusted Access for Cyber. Today we're a verified member of Anthropic's Cyber Verification Program. Both programs solve the same

Announcement

Escape joins OpenAI’s Trusted Access for Cyber (TAC) to advance AI-powered offensive security

We've been approved for OpenAI's Trusted Access for Cyber preview. For years we've been building toward one idea: offensive security that runs continuously inside engineering, instead of arriving twice a year as a PDF nobody reads past the executive summary. Business-logic DAST first, then

Product updates

Whitebox pentesting is now available in Escape's AI Pentesting - Cascade

Run Escape's AI Pentesting with your source code.

Research

Escape Research team found a PII disclosure in Keycloak. It's now CVE-2026-17059.

Keycloak filters its main users list so a restricted admin sees nothing. The endpoint that lists a role's members skips that filter and hands the same account everyone's email and name. Escape research found it, reported it, and it's now tracked as CVE-2026-17059.

Offensive Security

Automated Penetration Testing: The Complete Guide in 2026

Discover a guide to what automated penetration testing is, how it works, and best practices when looking for a vendor and implementing automated pentesting.

AI pentesting

7 best continuous penetration testing tools in 2026

Continuous penetration testing tools close the gap in coverage left by annual pentests. This guide breaks down the best options, what each tool actually does, and how to choose.

Case Study

How Escape DAST helped Sigma Computing achieve complete GraphQL API endpoint coverage

Sigma Computing needed GraphQL-native DAST, not a REST crawler with GraphQL bolted on. Here's how they got full endpoint coverage and slashed their triage time with Escape.

Research

Why does the rise of Agentic AI force security teams to become research-aware?

Bruce Schneier : "security is a process, not a product". It’s not enough to audit your application once, every new feature introduces new attack surfaces that must be taken into account when evaluating the security of the application as a whole. For example, a company rolling out a

Product updates

AI pentesting, Mission log: July

What we shipped in AI pentesting this month.

Product updates

wp2shell (CVE-2026-63030 + CVE-2026-60137): WordPress pre-auth RCE, now detected by Escape

wp2shell is an unauthenticated RCE chain in WordPress core. It combines two separate vulnerabilities: CVE-2026-63030 and CVE-2026-60137. Escape detects it across DAST and AI Pentesting, confirms exploitability, and shows affected assets within its Attack Surface Management.

DAST

Best AI DAST tools in 2026: ranked, compared, and reviewed for enterprise security teams

Every DAST vendor put "AI" on the homepage this year. The word now means ten different things to different tools, and even when discussing with analysts, they mention they give an "AI" badge when a certain capability checks the box. This is our attempt to fix

DAST

How Escape DAST Bypassed Immich's Locked Folder By Finding a Missing Default

How the Escape DAST uncovered a missing default that lets any Immich user read locked-folder photos without the PIN.

Application Security

Best GraphQL security tools in 2026: An in-depth guide, including business logic and enterprise coverage

Discover a practical guide to GraphQL security tools, an in-depth comparison, and how to choose the right ones for your stack.

Product updates

Fix issues with a prompt built for your coding agent

There's a translation step between a security finding and a fix, and it's usually done by hand. Someone opens the issue, reads the description, checks the CVSS score, follows the link to the affected asset, works out which endpoint is actually involved, then writes all of

Announcement

Introducing Escape's Partner Program

Join Escape's partner program to build offensive security into every stack, together.

AI pentesting

Modern AI-powered Pentesting Tools In-Depth benchmark

If you're evaluating AI for offensive security right now, you're weighing two questions at once. The first: why not skip the tooling and point a frontier model at your apps yourself? The second: among the AI pentesting tools you could actually buy, which one earns the

AI pentesting

Two Critical Vulnerabilities, One AI Pentester: How Cascade Found an Unauthenticated RCE and Walked Around the WAF

TL;DR We pointed Cascade, Escape's AI pentesting solution, at a single Spring + JSP customer portal. It came back with two findings that are typically difficult for traditional Dynamic Application Security Testing (DAST) scanners to detect: * Unauthenticated RCE via SpEL injection. A ref request parameter was dropped, unsanitized,

Pentesting

Race against the Machine: Can you match AI's pentesting time?

Race our AI in a series of pentesting challenges to try and beat its time.

Case Study

How Escape helped Bridgetech Group replace pentests with automated evidence

Security scanning that doesn't feel like going to the dentist. See how Bridgetech Group uses Escape for authenticated security testing and API scanning, replacing costly manual pentests with customer-ready security evidence for ISO 27001 and enterprise procurement.

Webinar recap

Claude in Security Engineering: From Proof-of-Concept to Production

Senior Security Engineer Geet Pradhan and Escape CEO Tristan Kalos run through how to use Claude to automate security engineering through threat modeling, detection engineering, and incident response

Escape - Application Security & Offensive Security Blog