Recognise risks
Make attack paths and trust boundaries visible.
From threat modeling to a simulated incident: learn where AI applications become vulnerable and put decisions to the test.
Themed afternoon · STARTPLATZ Düsseldorf · arrivals from 13:15 · sessions from 14:00
14:00–21:00 overall: themed afternoon and evening event
Make attack paths and trust boundaries visible.
Work through a ransomware attack together.
Examine access, permissions and safeguards.
How can security risks be identified before AI applications go live? Threat modeling through a practical agent example.
View sessionPhishing, overlooked vulnerabilities and failed backups: learn from real incidents how companies end up in crisis.
View sessionWhat happens when an AI agent can use internal data and tools? Understand the risks, then put them to the test in “Roast My Agent”.
View sessionExperience a simulated cyberattack, make decisions and understand the consequences.
View sessionAssess an agent test instance and identify suitable safeguards.
View sessionTalks and hands-on sessions complement each other. Stephan’s talk takes place before “Roast My Agent”.
More sessions to follow.
Welcome at 13:45. Exact times for individual sessions will be added.
Go to the evening event ↓Live demos, code and architecture. An evening for people building their own AI projects and sharing experiences with fellow developers.
Free · separate registration required.
Starts at 18:00 with arrivals and networking · official programme ends at 21:00.
You do not need a paid day pass. Register separately with AI Tinkerers. A day or week pass does not replace this registration.
Where does AI make a company vulnerable, and where does it protect it?
Which tools and data should an agent access? Where can an AI application be vulnerable? How should credentials and permissions be protected?
These three areas connect security principles to specific implementation questions.
Focused expert talks and guided application sessions run in parallel. Continue the work with specialist and technical support.
For your projectAnalyse a defined security question and identify the next checks or improvements it calls for.
Brigite Peposhi — What Could Go Wrong? Threat Modeling AI Before It Ships
Alexander Haunhorst — Three ways to disaster: how ransomware brings companies to their knees
Stephan Skrobisch — When your AI agent becomes an insider threat
Ransomware live with Alexander Haunhorst: from the first click to the last chance of recovery.
Roast My Agent with Stephan Skrobisch: examine an AI agent’s weaknesses and discuss protective measures. His talk takes place before the practical session.
Exact session times and practical-session durations will follow. Further contributions are welcome.
Sessions cover a defined scope; they do not constitute a full security audit of participants’ systems.
Day passes from €150 excl. VAT. The personal week pass brings all five days together.
Want to share your experience or run a hands-on session?
Help shape AI Week ↗