The "Explainable Machine Learning for Speech and Audio" workshop aims to enhance research in interpretability within audio and speech processing using neural networks. It addresses the challenges of making these neural network models more transparent, as their current "black box" nature affects trust and adoption in crucial areas like healthcare. The workshop will explore various explanation methods, from posthoc-explanations that provide human-understandable interpretations to models designed for inherent explainability, though sometimes at the cost of performance. Key discussions will revolve around real-life use cases, approaches that balance performance with interpretability, methods for generating explanations, evaluation of interpretation quality, and adapting interpretability techniques from other data modalities to audio. Submit your papers on Explainable AI for Speech and Audio for ICASSP 2024. Two submission tracks: 1. IEEEXplore for novel work (deadline: Jan 20, 2024) and 2. Workshop for works-in-progress or previously published papers (deadline: Feb 20, 2024). Best paper wins a prize!