AIF360, Fairlearn, and HELM — three open-source toolkits that catch unfair models. Measure fairness, mitigate bias, and benchmark LLMs across demographic groups.