Modern automotive audio systems have evolved into distributed, software-defined platforms that combine traditional DSP with machine learning. These systems are continuously trained, tuned, and updated using real-world data, creating complex workflows that extend beyond the vehicle. As a result, AI-driven audio technologies have become valuable intellectual property. Development and deployment typically occur across distributed environments. Engineers capture real-world data, refine models, and validate performance, while finalized algorithms are deployed onto embedded vehicle hardware for real-time operation. This lifecycle introduces significant security risks. Development tools may expose proprietary models outside controlled settings, and deployed systems are vulnerable to reverse engineering, extraction, or unauthorised reuse—especially when physical access to hardware is possible. This workshop focuses on safeguarding AI-driven audio across its lifecycle, identifying where IP is most exposed, and examining common attack methods. It also presents practical strategies to secure both development workflows and in-vehicle systems, ensuring innovation, collaboration, and long-term business value are protected.
