“Transparency requirements” and “content identification”: the Council of Europe adopted on Friday the first legally binding international treaty governing the use of artificial intelligence (AI).
“The treaty, also open for signature by non-European countries, establishes a legal framework that applies throughout the life cycle of AI systems, addresses the risks that these systems can represent and promotes responsible innovation,” specifies the European body in a press release.
The text was adopted at the annual ministerial meeting of foreign ministers from the 46 members of the Council of Europe.
This framework convention was developed over two years by an intergovernmental body which brought together the 46 member states, the European Union and 11 non-member states (the United States, Australia, Canada, Israel and Japan, notably), as well as representatives of the private sector, civil society and academia acting as observers.
“The convention sets out transparency and control requirements adapted to specific contexts and risks, in particular the identification of content generated by AI systems,” specifies the Council of Europe.
“The Parties shall adopt measures to identify, assess, prevent and mitigate possible risks and assess the need for a moratorium, ban or other appropriate measures regarding the use of AI systems where such use is likely to present risks incompatible with human rights standards,” continues the European body.
Marija Pejcinovic Buric, Secretary General of the Council of Europe, welcomed the adoption of a “unique international treaty that will ensure that artificial intelligence respects people’s rights”.
The agreement “responds to the need to have a standard of international law benefiting from the support of States from different continents united by common values, which makes it possible to take advantage of the advantages of artificial intelligence, while reducing the risks that it represents. With this new treaty, we intend to guarantee the responsible use of AI, respectful of human rights, democracy and the rule of law,” she added.
The framework agreement will be opened for signature in Vilnius on September 5 during a conference of justice ministers.
The European Union also adopted a law at the beginning of March to regulate AI systems like ChatGPT, a unique piece of legislation in the world.