AI as Defender: Threat Modelling and Security Testing
AI for security: useful, if you can trust it
Threat modelling, security testing, and keeping your risk assessment up to date take time that small software teams rarely have. Agentic AI workflows can take over a large part of that work, but only if you give them the right context and check their output. During this coaching day, you apply a multi-agent workflow to your own product, from security specification and threat analysis to targeted, AI-assisted pentesting. You learn where AI adds value, where it fails, and how to keep human review in control.
For whom?
Software developers and tech leads who want to use AI in their security practice
Security engineers and pentesters exploring agentic workflows
Product owners and CTOs responsible for risk assessment and CRA documentation
Participants of the Security by Design masterclass who want to take the next step
Key takeaways
This coaching day combines expert guidance, demonstrations, and group exercises. You will learn how to:
- Structure a security specification that gives AI the context it needs
- Use a multi-agent workflow to identify and assess threats
- Recognise unreliable AI output and organise human review
- Translate threat scenarios into targeted security tests
- Run AI-assisted pentests in a controlled test environment
- Keep your threat model and risk assessment up to date as your product evolves
You will leave with:
- A reviewed specification and diagram for a selected part of your product
- A short threat register with sources, assumptions, and prioritisation rationale
- 2 test cases linked to threats, with execution evidence from the test environment
- An updated control and residual-risk assessment
- A plan to continue the workflow in your development lifecycle
Programme
A full day that takes you from product context to tested, documented security risks.
| Time | Subject |
| 09:00 - 09:30 | Welcome and introduction • Objectives and participant challenges • The role of AI in security by design • Link with risk assessment and CRA documentation |
| 09:30 - 10:30 | Product context and security specifications • What AI needs: purpose, architecture, assets, and trust boundaries • Doomsday scenarios and existing controls • Human expertise, assumptions, and incomplete information |
| 10:30 - 10:45 | Coffee break |
| 10:45 - 12:00 | Agentic workflows for threat modelling • Specialised agents for specification, threat discovery, and risk reassessment • Threat knowledge sources, traceability, and human review gates • Guided exercises with practical examples |
| 12:00 - 12:30 | Quality and reliability of AI-generated analysis • Common pitfalls: wrong assumptions, irrelevant threats, and incomplete coverage • Benchmarking and expert review |
| 12:30 - 13:30 | Lunch break |
| 13:30 - 14:15 | From threat model to security testing • Linking threat scenarios and attack paths to pentesting • Test scope, prerequisites, and evidence requirements |
| 14:15 - 15:15 | AI-assisted pentesting in practice • Demonstrations and guided exercises in a controlled test environment • Agent-supported test preparation and interpretation of findings • Human supervision and operational limits |
| 15:15 - 15:30 | Coffee break |
| 15:30 - 16:15 | Continuous risk assessment and security evidence • Using test evidence to validate controls and residual risk • Updating the threat model after new vulnerabilities or changes • Versioning and accountability |
| 16:15 - 16:45 | Group coaching on your cases • Implementation challenges in your development practice • Practical tips for small teams with limited resources |
| 16:45 - 17:00 | Closing and next steps • Key lessons and questions • Next steps for implementation |
Registration
Seats are limited to allow personal coaching. To get the most out of the day, basic knowledge of threat modelling is recommended, for example through the masterclass Security by Design for Software Developers on 21 January 2027.
Flemish companies can take advantage of a subsidy from VLAIO for this course
This programme is part of Industriepartnershap in which 13 Flemish innovation partners offer an integrated service to stimulate growth and innovation in the Flemish industry in the 6 following themes: AI, Cybersecurity, Circularity, Digitisation, Climate & Energy and Industry 4.0. They do so under the leadership of Agoria and Sirris and with the support of Agentschap Innoveren & Ondernemen.
Date
Location
Price
Language
English

