UNESCO Recommendation on the Ethics of AI P5: paras 32 to 34 Privacy and data protection: handle data lawfully across its life, with multi-stakeholder governance and privacy impact assessment by design

Privacy is to be upheld at every stage of AI, and data used for AI should be handled, from collection through sharing and storage to deletion, in line with international law and the Recommendation. Adequate data protection frameworks with judicial protection should be set up, drawing on international principles on collection, use, disclosure and data subject rights, with a legitimate aim and valid legal basis including informed consent. Algorithmic systems need privacy impact assessments that include societal and ethical considerations and privacy by design, and AI actors are accountable for protecting personal information.