UNESCO Recommendation on the Ethics of Artificial Intelligence

The UNESCO RecommendationExternal link: sets out clear ethical guidelines and translates them into specific policy objectives. It thus establishes a legal basis under international law for initiatives at the level of member states, with the aim of ensuring that AI helps to respect, defend, and promote human rights and fundamental freedoms.

The Recommendation was developed by UNESCO following an intensive, two-year and occasionally controversial intergovernmental negotiation process. Numerous other stakeholders were involved in addition to the Member States themselves, including through an expert group and regional stakeholder consultations. The result is a global framework for the ethical use of AI, guiding the actions of all states regardless of their form of government or economic power. The global scope of the Recommendation is a key distinction from comparable existing initiatives, such as those of the OECD, the Council of Europe or the European Union.

Framework to protect human rights and fundamental freedoms

The UNESCO Recommendation sets out values and ethical principles, such as privacy, transparency, explainability and non-discrimination, that must be observed when developing and applying AI systems. The aim is to help shape the digital transformation in a way that promotes human rights and contributes to the United Nations' Sustainable Development Goals. To this end, the second part of the Recommendation sets out specific policy tasks and measures for the member states. These tasks relate to eleven policy areas, including the environment, education, science, communication, health and culture.

In addition to focusing on specific policy areas, the UNESCO Recommendation addresses the broader issue of ethical impact assessment and regulation. For instance, it suggests carrying out an Ethical Impact AssessmentExternal link: of AI systems. This tool is designed to identify and evaluate the potential benefits and risks to human rights and fundamental freedoms posed by AI systems throughout their entire life cycle. The Recommendation also states that AI systems should not only respect and protect human rights and fundamental freedoms, but actively promote them. Referring to the precautionary principle, it clearly defines the ethical framework for AI systems: if there is a reasonable possibility of negative consequences arising from the use of certain AI systems or technologies, their use or development should not be pursued. In concrete terms, this means putting an end to the use of AI for mass surveillance and social scoring systems.

New areas for policy action

The UNESCO Recommendation also goes beyond existing initiatives at the level of individual policy areas, identifying new areas where action is needed. In the area of social diversity, for example, it calls not only for measures against discrimination by AI systems, for example on the basis of gender or sexual orientation. It also encourages member states to promote diversity within the technology sector itself, where women, among others, remain significantly underrepresented. At the same time, they should disclose and combat any stereotyping in the results of AI systems and data. The UNESCO Recommendation also addresses the relationship between AI and environmental protection. In particular, it focuses on ensuring that AI systems are used in a sustainable and resource-efficient manner.

Implementation of the UNESCO Recommendation in Germany

The UNESCO Recommendation on the Ethics of AI provides a blueprint for the responsible development of national AI policies. Member States must now use this framework to establish processes and take measures to ensure the ethical use and development of AI in their countries. The German Commission for UNESCO is supporting the implementation of the Recommendation in Germany. It publicises the text and provides German ministries, parliamentsExternal link: and civil society organisations with information and advice on implementation options. To this end, the Commission organises specialist events, commissions scientific reports, and develops concrete courses of action in collaboration with experts.