New laws increasingly require large companies to check and report on human rights and environmental risks in their supply chains. Should companies also have to disclose publicly when and how they use AI for these checks? Should such AI tools be independently audited or certified?