January 13, 2026

Generative AI Security in the Automotive Domain

39 minutes

As generative AI (GenAI) shifts from traditional deterministic algorithms to complex probabilistic models, securing AI implementations across automotive systems presents entirely new engineering and cybersecurity challenges.

In this episode, host Zane Pelletier sits down with Ivan Granero, Cybersecurity Specialist at Bosch, to explore the security implications of deploying GenAI within the automotive domain. Ivan breaks down the fundamental shifts in risk modeling, the threat of prompt injection attacks, LLM data leakage, and how security teams must adapt threat analysis and risk assessment (TARA) frameworks to safeguard embedded automotive software and cloud-connected vehicle architectures.

Whether you are an automotive cybersecurity researcher, AI engineer, or vehicle software architect, this conversation provides a practical guide to securing next-generation AI in software-defined vehicles.

In this episode:
00:00 - Introduction: GenAI Security in Automotive
04:15 - Traditional Machine Learning vs. Probabilistic GenAI Models
10:30 - Emerging Threat Vectors: Prompt Injection & Data Leakage in LLMs
16:50 - Securing GenAI Across the Vehicle Lifecycle & Supply Chain
23:15 - Adapting TARA (Threat Analysis & Risk Assessment) for AI Systems
30:40 - Safeguarding Intellectual Property & Embedded Model Integrity
35:20 - Key Takeaways for Future Automotive AI Defense

To view the tool Zane and Ivan mentioned in their discussion, click here.

Thanks for listening!