Researchers proposed an AI-driven framework using Graph Neural Networks to optimize relay selection in NR-V2X systems, improving connectivity and reducing execution time by orders of magnitude. This could be relevant to developers building AI systems for smart city and Industry 4.0 applications.