Be it to detect social benefit fraud, to assign refugees to different regions, or to process tax declarations, the use of algorithmic or so-called AI-based systems in public administration is increasing at full speed. Recognizing this, the Council of Europe – the Strasbourg-based international organization with 46 member states and the mandate to protect human rights, democracy, and the rule of law – is introducing various frameworks to govern these automation processes. Most importantly, it is currently drafting a Convention on AI that should become binding international law.
In addition, the Committee on Legal Cooperation is focusing on guidance for public authorities in their use of algorithmic systems. To this end, it set up an expert group – called CDCJ-ADMIN-AI – which is tasked to elaborate proposals on the direction the committee should take in respect of future work in the area of administrative law and Artificial Intelligence. In addition, it shall review and update the Council of Europe Handbook «The Administration and You» in order to include any relevant aspects related to the use of algorithmic /AI-based systems in administrative law.
The working group held its inaugural meeting on 3 November, 2022, and is expected to conclude its work by December 2023. It is composed of representatives of member states, experts of independent organizations, and independent consultants. As its member, AlgorithmWatch will contribute to ensuring that the interests of individuals and the protection of their basic rights serves as the benchmarks for any automation of administrative processes. While algorithmic systems can certainly make sense in public administration, they must be deployed in a way that makes sure that they actually benefit those to which authorities have a unique responsibility – those subjected to it, that is, all of us.