AI Pentesting
AI-Powered Penetration Testing
We are developing an AI-driven penetration testing service that makes professional-grade security testing accessible and affordable, with the goal of offering it at roughly 10% of the cost of a conventional manual pentest. It is not available yet. Sign up now to join the beta programme in the second half of 2027, at a heavily reduced rate.
How it works
What is AI penetration testing?
AI pentesting combines automated AI-driven assessment with expert human validation. It is tailored to your industry and your technology stack rather than a one-size-fits-all scan. The service is delivered in three layers.
Continuous discovery
Guardian360 Lighthouse, our existing platform, continuously discovers your internet-facing and internal assets and keeps an up-to-date picture of your attack surface.
AI-driven assessment
Our own AI Cyber Security Agent generates and executes penetration-test scenarios tailored to the target industry and technology stack, going well beyond a generic vulnerability scan.
Expert review
A senior certified penetration tester, from Guardian360 or one of our selected partners, validates every AI finding before the final report is issued, so what you receive is expert-checked, not raw machine output.
Tailored, not generic
Scenarios are generated per industry and per technology stack, not a fixed checklist.
AI plus human
Automated assessment finds candidates at scale; a certified tester validates the findings that matter.
Around a tenth of the cost
Our goal is to offer professional-grade penetration testing at roughly 10% of the cost of a conventional manual pentest.
European and open source
Developed in Europe, with the agent, scenario libraries and methodology open-sourced back to the community at project close.
Scope
Penetration tests available in the beta
The beta programme will offer five types of penetration test, each generated and validated through the three-layer approach.
API penetration test
Testing of REST, GraphQL and other application programming interfaces for authentication, authorisation and logic flaws.
Cloud environment penetration test
Assessment of cloud configurations and exposure across your cloud estate.
Mobile application penetration test
Testing of Android and iOS applications and the services behind them.
Network penetration test (wired and wireless)
Assessment of internal and external network infrastructure, wired and wireless.
Web application penetration test
Testing of web applications against the OWASP Top 10 and beyond, with business-logic checks.
Who it is for
Professional pentesting, within reach
AI penetration testing is designed to put professional-grade testing in reach of organisations that could not previously afford it.
SMEs on a budget
Small and medium-sized enterprises that need real penetration testing but find conventional prices prohibitive.
Regulated sectors
Organisations in healthcare, government, finance and digital infrastructure that must comply with frameworks such as the NIS2 Directive, NEN 7510, the BIO framework or DORA.
Supply-chain partners
Suppliers to larger organisations who need to demonstrate their security posture as part of a supply chain.
The difference
Conventional vs AI penetration testing
A conventional manual penetration test is thorough but slow and expensive. AI pentesting keeps the expert validation and makes the rest faster and far cheaper.
| Conventional manual pentest | Guardian360 AI pentesting | |
|---|---|---|
| Indicative cost | From EUR 6,000 to well over EUR 100,000 | Our goal: around a tenth of that |
| Turnaround | Weeks, subject to tester availability | Faster, AI-generated and expert-validated |
| Tailoring | Depends on the individual tester | Generated per industry and tech stack |
| Human validation | Yes | Yes, senior certified tester reviews every finding |
| Accessibility | Out of reach for many SMEs | Designed for SMEs and essential entities |
Open source
Given back to the European community
At project close in March 2028, Guardian360 will open-source significant parts of the project, a deliberate contribution to the European open-source cybersecurity ecosystem and a condition of the EU funding.
The AI Cyber Security Agent
The engine that generates and executes penetration-test scenarios.
Industry-specific scenario libraries
Tailored to verticals such as healthcare (NEN 7510), government (the BIO framework) and finance (DORA).
Methodology and documentation
So others can build on, audit and improve the approach.
We believe professional-grade security tooling should be accessible, not just as a service, but as a foundation others can build on.
Jan Martijn Broekhof, founder and CEO of Guardian360
Beta programme
Become a beta tester
Beta testers get early access to the AI pentesting service in the second half of 2027, and help shape it.
What you get
- Early access to the AI pentesting service in H2 2027.
- A heavily reduced rate compared to the eventual commercial price.
- A full pentest report with expert-validated findings and remediation guidance.
- The opportunity to shape the service with your feedback.
What we ask
- Your organisation is willing to provide access to the agreed scope for testing.
- The person signing up is authorised to request a penetration test.
- Results and methodology may be used, anonymised, to research and improve the service.
Sign up
Join the beta programme
Leave your details and we will be in touch when the beta programme opens in the second half of 2027. Signing up is not a commitment.
FAQ
Frequently asked questions
What is AI penetration testing?
AI penetration testing combines automated, AI-driven assessment with expert human validation. Our own AI Cyber Security Agent generates and executes pentest scenarios tailored to your industry and technology stack, and a senior certified penetration tester validates every finding before the report is issued. The AI augments the expert; it does not replace them.
How much does it cost?
Our goal is to offer professional-grade penetration testing at roughly 10% of the cost of a conventional manual pentest, which can run from EUR 6,000 to well over EUR 100,000. Beta testers get an additional, heavily reduced rate.
When will the service be available?
The service is still in development. The beta programme is scheduled for the second half of 2027, with general availability expected in 2028. The project runs from September 2026 to March 2028.
What types of penetration test are supported?
Five types: API, cloud environment, mobile application, network (wired and wireless), and web application penetration testing.
Who is eligible for the beta programme?
Any organisation in the European Union can apply. It is particularly relevant for SMEs, entities in scope of the NIS2 Directive, and organisations in healthcare, government, finance and digital infrastructure, as well as supply-chain partners who need to demonstrate their security posture.
Is the technology open source?
Yes. At project close in March 2028, Guardian360 will open-source the AI Cyber Security Agent, the industry-specific scenario libraries and the methodology and documentation, contributing back to the European cybersecurity community.
Who funds this project?
The project is funded by the European Union under the CYSSDE programme, an initiative supported by the European Cybersecurity Competence Centre (ECCC). Guardian360 received a sub-grant under CYSSDE Open Call 3, and will deliver 150 AI-assisted penetration tests across multiple verticals during the 18-month project (September 2026 to March 2028).
Co-funded by the European Union
This project has received funding from the European Union under the CYSSDE programme (Grant Agreement No. 101172498). Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or ECCC. Neither the European Union nor the granting authority can be held responsible for them.