CAI's Terms of Reference 

 Revised "Zero Draft"

At its 4th Plenary meeting, the CAI decided to make the revised “Zero Draft” [Framework] Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law public.

This document was prepared by the Chair of the CAI with the support of the Secretariat to serve as the basis for the drafting of the [Framework] Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law. It does not reflect the final outcome of negotiations in the Committee.

 Consolidated Working Draft

Following the first reading of the revised Zero Draft, the Chair of the CAI prepared, with the assistance of the Secretariat, a Consolidated Working Draft to serve as the basis for further negotiations of the Framework Convention, containing provisions which have preliminarily been agreed to during the first reading of the revised Zero Draft, as well as proposals drafted by the Chair with the assistance of the Secretariat.

The CAI Bureau decided at its 6^th meeting to make this document public. It does not preclude the final outcome of negotiations in the CAI.

Artificial Intelligence (AI) – the very phrase conjures up visions of science fiction with robots and all-powerful computers running the world for humans in some distant future. Yet AI systems are already pervading our daily life, from our   smart phones over social media to self-driving cars. Such systems are also being   used for various administrative tasks by public authorities in many countries. In  addition, the rapid spread of “general purpose AI systems” has presented us with further new opportunities, as well as new challenges.

The question is therefore no longer if we want to make use of these powerful tools, but how we ensure that they are used for the good of humanity only.

Given the fact that AI is developed and used across borders, it makes sense to develop a common global approach to basic principles which should govern how we, as humanity, develop and use AI systems. 

The Committee of Ministers has tasked the Committee on Artificial Intelligence (CAI) with elaborating a legally binding instrument on the development, design and application of AI systems based on the Council of Europe’s standards on human rights, democracy and the rule of law, based on such basic principles. At the same time, the instrument shall be conducive to innovation.

The focus of the [Framework] Convention will be on ensuring the continued seamless application of human rights and the principle of rule of law in contexts where AI systems assist or replace human decision-making or perform other tasks relevant in such contexts. Moreover, AI systems shall only be used in such a way that they do not, directly or indirectly, endanger or undermine democratic processes.

To develop such a [Framework] Convention as a global instrument is a challenging task. We all agree that we intend to develop an instrument that is attractive, not only to States in Europe, but to as many States as possible from all regions of the world. The more global the instrument becomes, the more impact it will have on peoples’ lives around the globe. We therefore call on all States of the world that uphold the values of human rights, democracy and the rule of law to join our endeavours to develop a set of shared basic principles applicable to the design, development and application of AI. These principles should be strong and clear, but formulated in such a way that all Parties retain the necessary margin of appreciation to implement them in the logic and culture of their legal systems and traditions. We also need to take into account that, while some States and regional organisations are in the final steps of developing domestic legislation, others are still in the process of doing so.

We are also committed to including the views and experiences from all stakeholders including intergovernmental organisations, academia, businesses and civil society, in their respective roles, into our work.

Our working methods allow all observers to actively participate in the negotiations on the draft [Framework] Convention and meaningfully contribute to the elaboration of the legally binding instrument.

Finally, we must also be aware that the instrument we are currently working on cannot regulate all aspects of the development and use of AI systems. Further binding and non-binding instruments, both regarding specific aspects of AI and its use in specific sectors, will be needed to comprehensively address the use of this complex and rapidly evolving technology.

Given the immense societal importance of the topic of AI and how this technology affects human rights, democracy, and the rule of law, I wish to ensure as open and transparent a process as is possible in international negotiations and will therefore on a regular basis provide public updates on how our work progresses.

I look forward to sharing this fascinating and exciting journey with you all!  

Thomas Schneider (Switzerland)
Chair of the CAI