Following a second session of an ad hoc expert group meeting focusing on the ethical aspects of Artificial Intelligence in early September 2020, UNESCO released a draft recommendation on 7 September 2020. It aims at providing a framework for developments in artificial intelligence. This draft concludes the process launched in mid 2020 after the first session of the expert group and a submission for public contribution. The document is a first step to prepare future discussions between Member States which should occur in 2021. The framework envisaged in this document is based on a set of values and principles that should be implemented in different policy areas.
The four values put forward in the draft are following those promoted by UNESCO:
- Respect, protection and promotion of human dignity, human rights and fundamental freedoms.
- Environment and ecosystem flourishing.
- Ensuring diversity and inclusiveness.
- Living in harmony and peace.
In order to complete and clarify these values, the document sets out 10 principles to be applied to the use of artificial intelligence:
- Proportionality and do no harm.
- Safety and security.
- Fairness and non-discrimination.
- Human oversight and determination.
- Transparency and explainability.
- Responsibility and accountability.
- Awareness and literacy.
- Multi-stakeholder and adaptive governance and collaboration.
Finally, UNESCO identifies 10 policy areas in which these values and principles must find an operative translation:
- Ethical Impact Assessment.
- Ethical Governance and Stewardship.
- Data Policy.
- Development and International Cooperation.
- Environment and Ecosystems.
- Education and Research.
- Economy and Labour.
- Health and Social Well-Being.
In all these policy areas, the document encourages Member States to adopt legislation and procedures that would make the set-out values and principles effective. UNESCO envisions Artificial intelligence in a human-centric perspective focusing on the human being. The document therefore develops a framework to use AI as a vector tool for reducing inequalities, protecting the environment and human rights as well as access to education and culture. UNESCO sees these technological tools as an additional vector for achieving the organisation’s goals and not as a technology that would overturn major established legal principles. The transposition of the substance of this document into practice should come from Member States and their willingness to adopt a final recommendation in line with that draft.