Background: In 2022, there were 18 weather and climate disaster events in the U.S. designated by NOAA as "billion-dollar disasters." These events caused hundreds of deaths and financial losses exceeding $165 billion. They included 11 severe storms, 3 hurricanes, 1 flood, and 1 wildfire. Traditional environmental modeling of severe storms, hurricanes, floods, and wildfires can often be too costly, too slow for decision makers, and face computational limitations. Applying new technologies, such as artificial intelligence (AI), to models of these hazards can provide significant improvements in forecast skill, accuracy, and coverage. However, applying AI to these models still has challenges and limitations, including technical and regulatory issues. For example, to effectively train models, AI requires large amounts of data, which are often lacking for these hazards. Furthermore, inadequate or otherwise flawed data can produce biased results. Key Questions: (1) How, if at all, has the application of AI affected modeling capabilities for severe storms, hurricanes, floods, and wildfires? (2) How, if at all, could modeling capabilities for severe storms, hurricanes, floods, and wildfires be affected by future applications of AI? (3) What challenges exist with regard to the development and application of AI in modeling capabilities for severe storms, hurricanes, floods, and wildfires? (4) What policy options might help address the potential challenges related to the development and application of AI in modeling capabilities for severe storms, hurricanes, floods, and wildfires?