The field of autonomous driving and explainable AI is rapidly advancing with a focus on developing innovative techniques for collision avoidance and model interpretability. Recent research has led to the creation of closed-loop frameworks that integrate risk assessment with active avoidance control, enabling proactive and effective collision avoidance in dynamic interactive traffic. Additionally, novel approaches have been introduced to learn collision risk from naturalistic driving data without requiring labor-intensive annotation of sparse risk, capturing the patterns of normal driving and estimating the extent to which a traffic interaction deviates from the norm towards unsafe extremes. Furthermore, explainable AI methods have been proposed to provide human-interpretable explanations for machine learning models, including techniques for explaining neural networks and embedded vector spaces. These advances have the potential to significantly improve the safety and reliability of autonomous systems. Noteworthy papers include: REACT, which achieves 100% safe avoidance with zero false alarms or missed detections, and Learning Collision Risk from Naturalistic Driving, which establishes a scalable and generalisable foundation to proactively quantify collision risk in traffic interactions. Other notable works include Explaining Neural Networks with Reasons and Explainable embeddings with Distance Explainer, which provide innovative methods for model interpretability.