Artificial intelligence is advancing rapidly in wildfire science, yet significant challenges remain in meeting expectations that AI can accurately solve complex problems involving wildfire occurrence, burned area, fire behavior, emissions, smoke, air quality, and ecosystem impacts under sparse, uncertain, and rapidly changing conditions. This session invites contributions examining what AI can realistically achieve, where limitations remain, and what practical advances are emerging. Topics include not limited to AI for burned area, fire radiative power, emissions, smoke and air quality, hybrid physics-AI, uncertainty, and noisy or incomplete observations.

We also welcome contributions on trustworthy and usable AI, including decision support, human-AI interaction, foundation models, and agentic systems. We particularly encourage work addressing the gap between expectations and reality in AI performance, how AI can complement physically grounded wildfire science, and where AI is genuinely advancing discovery, prediction, and practical understanding across the broader wildfire research community.

• A - Atmospheric Sciences
• GC - Global Environmental Change
• GH - GeoHealth
• IN - Informatics