Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law
In recent months, the Council of Europe (“Council”) has reached an agreement on the Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law[1] (“AI Framework Convention”), which has been negotiated since 2021, and the adoption of the draft text was expected.
The Committee of Ministers of the Council adopted the AI Framework Convention on 17 May 2024 in Strasbourg.
The AI Framework Convention is the first internationally binding treaty aiming to ensure respect for human rights, the rule of law and the legal standards of democracy in the use of artificial intelligence systems.
The AI Framework Convention, which aims to establish some standards for all countries of the world in the context of artificial intelligence studies, draws attention with its openness to signature for countries outside Europe in support of this attitude.
The AI Framework Convention covers all the processes of artificial intelligence systems from the design stage to their launch, that is, their life cycle. As a matter of fact, the AI Framework Convention aims to be inclusive for all processes of artificial intelligence systems from the beginning to the end and to respect human rights, democracy and the rule of law.
The definition of artificial intelligence systems in the AI Framework Convention is in line with the Artificial Intelligence Act[2]; it is defined as a machine-based system that, for explicit or implicit purposes, extracts from its input how to produce outputs such as predictions, content, recommendations or decisions that may affect physical or virtual environments.
The first paragraph of the third article regulating the scope of the AI Framework Convention states that the Convention shall apply to public authorities or the private sector acting on their behalf and to the independent private sector.
However, States Parties are required to make a declaration on how they intend to implement the obligations of the AI Framework Convention, and it is made possible for them to change these declarations at any time.
AI Framework Convention contains general principles that States Parties should adopt and implement in their domestic laws. These principles are; respect for human dignity and individual autonomy, transparency and compliance with oversight requirements, accountability and responsibility, right to equality and non-discrimination, confidentiality and protection of personal data, ensuring reliability and safe innovation.
Accordingly, States Parties are required to take measures to identify, assess, prevent and mitigate potential risks and to implement a limitation or other mechanism for the use of artificial intelligence systems in cases where their risks may threaten a violation of human rights.
They will also have to provide accountability and responsibility for adverse impacts and ensure that AI systems respect equality, including gender equality, non-discrimination and privacy rights.
Addressing the risks that AI systems may pose to democracy requires States Parties to take measures to ensure that AI systems recognise and are not used to undermine democratic institutions and processes, including the principle of separation of powers, respect for the independence of the judiciary and access to justice.
It is envisaged that a follow-up mechanism will be established and international co-operation will be ensured to ensure effective implementation of the Convention by the State Parties. In addition, according to Article 23 of the Convention, a mechanism called "Conference of the Parties" consisting of representatives of the States Parties will be established.
The AI Framework Convention will be opened for signature on 5 September 2024 in Vilnius/Lithuania.
[1] AI Framework Convention by Comittee on AI, Council of Europe , May 17, 2024
[2] Artificial Intelligence Act by Europaen Parliament,March 13, 2024