Part of the Hi! PARIS "Fairness in AI" series, this session explores demographic parity in regression and classification under the unawareness framework, focusing on optimal fair predictors and transport costs.