Modern vehicles are highly complex systems that challenge automotive manufacturers in quality assuring their products. Until today, quality attributes (e.g., weld points, clamping locations) are often planned based on the expertise of engineers. Although the rise of software in the automotive domain has introduced new quality-assurance and planning tools like computer-aided design software or simulations, such software is often fragmented, difficult to use, and rarely standardized across product platforms. In parallel, artificial intelligence (AI) has become more capable and offers means to automate tasks, harmonize data structures, and generate reproducible recommendations. Still, it has not been explored to what extent AI-based software supports or challenges quality planning in the automotive domain. To tackle this gap, we report an interview survey with 13 experts in manufacturing, quality assurance, and software development from Volkswagen. Following qualitative research methods, we applied structured coding to analyze the interviews. Our results indicate persistent shortcomings in existing software tools, particularly with respect to usability, data integration, and reproducibility. At the same time, our interviewees pointed to promising AI use cases, including automated positioning of quality attributes, data validation, and simulations based on digital twinning. However, concerns remain about the trust, transparency, and validation of AI-generated outputs. Our study provides an overview of current practices in quality planning in automotive practice, highlighting factors to improve it through AI. So, we contribute insights for practitioners seeking to improve quality planning for complex systems, while researchers can identify opportunities to design AI-based quality planning for industry.