Researchers have proposed a new risk measure called the Wasserstein entropic value-at-risk. This measure is designed to capture an agent's uncertainty in its environment and provide a more robust way of assessing risk. Unlike traditional measures, it can account for potential catastrophes that were previously ignored. The authors derive a closed-form expression for a dynamic-programming operator that incorporates this new risk measure. --- Why it matters: This research matters because it provides a more accurate way to assess risk in uncertain environments, which is crucial for developing safe and reliable AI systems. Source: https://arxiv.org/abs/2608.19073