Talks
Presentations I have given in the past
Defenders of the Realm: Breaking and Fixing Keycloak Configurations
Keycloak Dev Dev 2026, Darmstadt · Recording
Also given at Keyconf 2025, Amsterdam · Recording
In this joint talk with my colleague Tim Walter, we introduce kcwarden, our Open Source Keyloak configuration security auditor.
As Keycloak practitioners, we often focus on implementing features rather than continuously evaluating our security posture. After being responsible for the security of a Keycloak instance in a large-scale project for nearly three years, we’ve encountered several configuration vulnerabilities that can easily slip through even experienced teams.
In this talk, we’ll go beyond the basics to explore common security pitfalls in Keycloak deployments that we’ve encountered during our journey. More importantly, we’ll introduce our open source tool, kcwarden, which we developed to automate security auditing of Keycloak configurations. This tool not only detects standard security issues but can be customized to identify organization-specific concerns such as problematic role assignments or policy violations, enabling continuous monitoring of your Keycloak environment.
Join us to discover how kcwarden can enhance your existing Keycloak deployment’s security posture and learn practical strategies for implementing automated configuration checks into your operational workflows.
Why Defensive Software Architecture is Important, or: How to Compromise a Payment Processor using Math
Elbsides 2025, Hamburg · No recording available
This talk discusses the same contents that I describe in this blog post. There is no video available, as the organizers had to unpublish their entire conference archive due to a copyright complaint :(.
People talk a lot about defensive software architecture, but does it really make a difference?
In this presentation, I give an example from a security audit we performed, in which a simple misuse of a cryptographic primitive led to the compromise of an entire customer service tool for a payment provider, which would have allowed us full administrative access to their backends.
By stepping through the architecture of the system from the highest level down to the exact vulnerable code, this example allows us to illustrate the advantages of defensive software architectures with multiple layers of security. By the end of the presentation, you will have a new appreciation for defensive software architectures. As a bonus, you will also have learned about a neat cryptographic trick that exploits unauthenticated encryption.
Protectors of the Realm: Breaking and Fixing Keycloak Configurations
Elbsides 2024, Hamburg · No recording available
This was an earlier version of the talk Tim Walter and I gave at Keyconf 2025. There is no video available, as the organizers had to unpublish their entire conference archive due to a copyright complaint :(.
Keycloak is a widely used tool for authentication and user management in many projects and is therefore often encountered in security audits. However, analyzing the security of an existing Keycloak instance is often a challenge due to the complexity of the software and its configuration options.
Having been responsible for the security of a Keycloak instance in a large software project for almost two years, we share our insights on what to look for. We also present our open source Keycloak configuration auditor kcwarden, which we developed to facilitate our work. The tool can detect common security issues and can be easily customized to detect project-specific issues, such as dangerous role assignments or policy violations. This allows us to continuously monitor the configuration for dangerous changes that might otherwise go unnoticed.
After attending this talk, you should have an overview of common Keycloak misconfigurations and how you can use kcwarden to support your work.
Wie man mit Mathematik ein API übernehmen kann (und wie gute Architektur das verhindert)
Gulaschprogrammiernacht 2023, Karlsruhe · Recording
A German talk discussing the same contents that I describe in this blog post.
Beim Pentesten stolpert man manchmal über Fehler, die viel zu viel Spaß machen, um sie nicht mit anderen zu teilen. So auch hier: in diesem Vortrag berichte ich von einer Sicherheitslücke, in der durch die falsche Verwendung von Kryptographie das Stehlen von Zugangsdaten für ein API Gateway möglich war - und wir damit Kontrolle über große Teile einer Finanz-Anwendung erhalten konnten. Dabei gehen wir durch die gesamte Schwachstelle bis herunter zur Bit-Ebene, um zu verstehen, wie der Angriff funktioniert.
Doch wir wollen auch noch einmal herauszoomen und uns anschauen, warum die Architektur des Systems die Sicherheitslücke so katastrophal gemacht hat, und wie eine defensivere Architektur an vielen Stellen den Angriff hätte verhindern oder eingrenzen können. Am Ende werdet ihr also nicht nur eine schöne Eigenheit eines verbreiteten Verschlüsselungsalgorithmus verstanden haben, sondern auch ein weiteres Beispiel haben, warum defensives Programmieren eine gute Idee ist.
Es ist kein Vorwissen über Kryptographie erforderlich, alles wird im Vortrag erklärt. Grundlegendes Wissen über bool’sche Logik ist zum Verständnis hilreich.
Responsible Disclosure at Scale - Ein Erfahrungsbericht
German OWASP Day 2023, Frankfurt · Recording
A German lightning talk I gave about a responsible disclosure research project I conducted as part of my PhD. Below is the German-language abstract. There is also an english-language version of the full conference talk on that topic.
In Zeiten von automatischen Scannern wird es zunehmend leichter, im großen Stil Webseiten auf Sicherheits- und Datenschutzprobleme abzuklopfen. Doch was macht man eigentlich, wenn man fündig geworden ist? Wie kriegt man die Betreiber:innen der Webseiten dazu, die Probleme zu beheben? Das war die Frage, die sich uns vor ein paar Jahren stellte - und wenn man schon an einer Uni promoviert, dann macht man eben ein paar Studien zu dem Thema.
In diesem Talk möchte ich euch einen kurzen Überblick darüber geben, was danach geschehen ist. Ich berichte von den Antworten, die wir auf unsere Benachrichtigungen bekamen - von Geschenken über Misstrauen bis hin zu Verschwörungstheorien. Und ich berichte über die Ergebnisse der Studien: Was hat funktioniert, und was nicht? Am Ende des Vortrags habt ihr hoffentlich eine bessere Idee, was euch bei einer solchen Aktion erwartet - im guten wie im schlechten.
Effective Notification Campaigns on the Web - A Matter of Trust, Framing and Support
USENIX Security 2021, Online · Recording
A talk describing a joint research project with Alina Stöver, TU Darmstadt; Henning Pridöhl, Universität Bamberg; Sebastian Bretthauer, Goethe-Universität Frankfurt; Dominik Herrmann, Universität Bamberg; Matthias Hollick, TU Darmstadt; Indra Spiecker, Goethe-Universität Frankfurt. Abstract below.
Misconfigurations and outdated software are a major cause of compromised websites and data leaks. Past research has proposed and evaluated sending automated security notifications to the operators of misconfigured websites, but encountered issues with reachability, mistrust, and a perceived lack of importance. In this paper, we seek to understand the determinants of effective notifications. We identify a data protection misconfiguration that affects 12.7 % of the 1.3 million websites we scanned and opens them up to legal liability. Using a subset of 4754 websites, we conduct a multivariate randomized controlled notification experiment, evaluating contact medium, sender, and framing of the message. We also include a link to a public web-based self-service tool that is run by us in disguise and conduct an anonymous survey of the notified website owners (N=477) to understand their perspective.
We find that framing a misconfiguration as a problem of legal compliance can increase remediation rates, especially when the notification is sent as a letter from a legal research group, achieving remediation rates of 76.3 % compared to 33.9 % for emails sent by computer science researchers warning about a privacy issue. Across all groups, 56.6 % of notified owners remediated the issue, compared to 9.2 % in the control group. In conclusion, we present factors that lead website owners to trust a notification, show what framing of the notification brings them into action, and how they can be supported in remediating the issue.