AI is accelerating software development, but it's also empowering attackers. For automotive ECUs, this requires a shift from isolated testing to a fully integrated security lifecycle. This workshop explores how AI-assisted techniques can be applied across ECU design, validation, and continuous improvement. First, we focus on AI-supported reverse engineering of ECU firmware.

• Participants will learn how AI tooling helps identify code structures, understand communication interfaces, and accelerate the analysis of unknown firmware, providing insight into the risks of AI-enhanced attacks.
• Next, we demonstrate how this intelligence drives an ontology-based security workflow. An AI penetration test agent leverages a central cybersecurity ontology—the ECU’s "security brain"—which holds design artifacts like TARA, security concepts, and S-BOM data. This allows the agent to create highly precise, targeted test cases based on predicted weaknesses, moving far beyond generic checks.
• Finally, we explore the critical feedback loop. Test results are fed back into the ontology, enabling a dynamic assessment of the ECU's security posture. This allows for powerful analysis: Has the likelihood of a TARA-defined attack path changed? Is a known vulnerability truly exploitable? This continuous cycle provides concrete evidence to refine security concepts and drive a more resilient ECU design.

This workshop is intended for security engineers, penetration testers, embedded developers, and automotive security professionals who want to understand how to build an AI-driven, end-to-end validation workflow that is deeply integrated with the cybersecurity development lifecycle.

Referees: 

Dr. Nils Weiß, Co-Founder (Dissecto GmbH)

Thomas Irmscher, Senior Product Manager (ETAS GmbH)

Discover how the Automotive Cybersecurity Trend Radar helps identify and prioritize emerging cybersecurity challenges and opportunities. This interactive workshop introduces the Trend Radar methodology, including how trends are identified, clustered, and assessed based on impact and maturity. Participants will actively contribute by reviewing existing trends, discussing new topics, and exploring how the Trend Radar supports roadmap planning, risk analysis, and strategic decision-making in automotive cybersecurity. As an addition, we will present a case study that investigates how AI and in particular vulnerability finding with AI changes the threat landscape.

Referees: 

Gabriel Simmann, Research associate (Karlsruhe University of Applied Sciences)

Frank Kargl, Professor (University Ulm)