Stanford researchers evaluate whether major foundation model providers comply with the proposed EU law on AI. They provide recommendations that policymakers prioritize transparency and providers take action to collectively set industry standards that improve transparency.