CAI's Terms of Reference 

The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law was adopted on 17 May 2024 by the Committee of Ministers of the Council of Europe at its 133th Session held in Strasbourg, and will be opened for signature on the occasion of the Conference of Ministers of Justice in Vilnius (Lithuania) on 5 September 2024.

  Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law

  Explanatory Report



With the formal adoption by the Committee of Ministers, the Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law negotiated by the CAI is now definitely the first binding international treaty on AI and waiting to be signed and ratified by countries.

In contrast to hopes and fears to the contrary, the negotiating parties have neither intended to create new substantive human rights nor to undermine the scope and content of the existing applicable protections. The intention of the parties negotiating the instrument has been to make sure that each party’s existing protection levels of human rights, democracy and rule of law would also apply to current and future challenges raised by AI.

From the very beginning of the process, all European and non-European States participating in the negotiations underlined the importance of the treaty being set up as a global instrument. In addition to the 46 Council of Europe member States, a number of countries from several regions (Argentina, Australia, Canada, Costa Rica, the Holy See, Israel, Japan, Mexico, Peru, Uruguay and the Unites States) participated in the negotiations, along with the European Union. In addition, the process was set up in an inclusive manner with many CoE bodies, other IGOs such as the OECD, OSCE, UNESCO and around 70 representatives from civil society, business and the technical and academic community actively participating and using the ability to make comments and text proposals to the draft treaty until the very last day of the negotiations.

To negotiate a binding instrument in such an inclusive process and within a very short deadline has been intense and sometimes challenging. We had to find ways to bridge differences between States’ legal systems and traditions including some difference in the interpretation of some human rights. We also had to manage a range of expectations about how to develop a legal instrument which would have an impact on AI governance at global level. After many days and nights of intense negotiations, all governments involved, representing 57 States, were able to agree on the result.

So this is it:

The Framework Convention formulates fundamental principles and rules which not only safeguard human rights, democracy and the rule of law but at the same time are conducive to progress and technological innovations. It is complementary to the already existing international human rights, democracy and rule of law standards and aims at filling-in any legal gaps that may have formed as a result of rapid technological advances in the sphere of human rights law but also with regards to the protection of democracy. Given the high level at which it is operating and in order to remain future-proof, the Framework Convention does not regulate technology and is essentially technology neutral. The Framework Convention and its implementation should follow the logic of a graduated and differentiated approach, in view of the severity and probability of adverse impacts on human rights, democracy and the rule of law.

The Framework Convention applies to both public and private sectors. Limited exemptions from the scope are foreseen as regards national security and research and development. In line with the Statute of the Council of Europe, matters relating to national defence do not fall within the scope of the Convention.

The Framework Convention puts an obligation on all future Parties to address the risks posed by activities within the lifecycle of AI by public and private actors, while taking into account the respective roles and responsibilities of all stakeholders and giving flexibility to Parties in meeting the agreed obligations according to their domestic legal and institutional frameworks and their respective international human rights obligations.

The treaty and its implementation framework will also provide for new opportunities for cooperation with all stakeholders, including with States which will not or not yet have ratified it. This will further contribute to its potential to have a global reach.

In order to adequately govern AI in the present and the future, our societies will need to develop a broad set of technical, legal and socio-cultural norms adequate to the various uses of AI in different societal and economic contexts in various regions of the world – comparable to what our societies have developed in the past two centuries in order to cope with the use of engines in different vehicles and machines used for various purposes.

So, the Framework Convention will not be a stand-alone instrument, but one important step in the development of a governance framework for AI, so that all people in our societies can benefit from AI systems and be part of innovative societies and economies, while maintaining and strengthening existing protections of human rights, democracy and the rule of law.

I am looking forward to the continuation of the work of the CAI, under a new Chair to be elected at the next Plenary meeting in September, on a methodology for the risk and impact assessment of AI systems on human rights, democracy and rule of law – which will be another milestone instrument complementing and operationalizing the Framework Convention, and other legal instruments.

I would like to deeply thank all those representatives from all stakeholder groups who have, despite many difficult moments, never given up to work hard and constructively, and with respect to each other’s roles and responsibilities, to allow us to reach consensus.

It has been a great honour and privilege to serve all of you in the past two years.

As ever,

Thomas Schneider (Switzerland)
Chair of the CAI from 2022 to 2024


Towards an application of AI based on human rights, the rule of law and democracy



 Contact us