a) AI actors should respect the rule of law, human rights and democratic values, throughout the AI system lifecycle.

b) To this end, AI actors should implement mechanisms and safeguards, such as capacity for human determination, that are appropriate to the context and consistent with the state of art.