insAIde #50 : the first global AI treaty to protect human rights is a great achievement
🌎 🤖 ✍️
It can now truly be said that the Council of Europe has adopted the first legally binding international treaty on artificial intelligence. It is the Council of Europe Framework Convention on artificial intelligence and human rights, democracy, and the rule of law. Approved a few months ago, it was signed a few days ago by the European Commission on behalf of the EU, as well as by the US, Israel, the UK and other countries.
The importance of the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law
The news did not have worthy prominence when the treaty was adopted, and so far, even the signing by the EU Commission does not seem to have aroused the interest it deserves. However, Padre Paolo Benanti spoke about it in Cernobbio with the right timing, alerting that qualified audience to the importance of this event, which, to be successful, requires the commitment of States, businesses, and citizens.
It is the first international treaty on AI
AI regulation should not remain relegated to the national or regional level but should become the subject of the broadest international reflection and sharing, thus paving the way for the conclusion of one or more international treaties.
The Framework Convention on Artificial Intelligence is the result of two years' work carried out by the Committee on Artificial Intelligence (CAI), an intergovernmental body set up by the Council of Europe to create a rule of international law shared by different states, united by common values.
The CAI brought together the forty-six member states of the Council of Europe, i.e. the countries of the European Union, the non-EU European countries (from the United Kingdom to the Vatican City State) and ten other states that are geographically not part of Europe (Argentina, Australia, Canada, Costa Rica, Japan, Israel, Mexico, Peru, the United States, Uruguay), as well as representatives from the private world, civil society and academia, to draw up the treaty adopted last May in Strasbourg.
The Framework Convention, accompanied by an interesting explanatory report, was opened for signature in Vilnius on 5 September last, and it was on that occasion that the European Union signed it. But it is worth remembering that the treaty is also open to accession by other non-European countries.
Inside the AI Framework Convention
The treaty consists of a preamble and thirty-six articles and aims to ensure that activities within the life cycle of artificial intelligence systems respect human rights, democracy, and the rule of law. It therefore requires each signatory party to adopt or maintain in force legislative, administrative, or other measures that are appropriate to implement the Convention. In this way, the treaty aims to address the risks that AI systems may present while promoting anthropocentric and responsible innovation.
By adopting the OECD definition of an AI system (and already used as a benchmark for the Artificial Intelligence Act), the treaty specifically addresses the use of AI systems in the public sector, while leaving it up to the signatories to decide how to regulate the life-cycle activities of AI systems by private entities in a manner consistent with the Convention.
The Convention thus lays down general obligations regarding the protection of human rights and the integrity of democratic processes and respect for the rule of law and then dictates a series of common general principles relating to the life-cycle activities of AI systems. They speak, for instance, of human dignity, transparency, accountability, non-discrimination, reliability, and protection of personal data.
The ethical key enters the law, permeates it, recalls it, and reconciles parts and aspects of it that are already regulated, through a common thread that is given by the indirect but constant reference to the principles and values of the European Constitution.
The treaty also includes provisions on remedies for human rights violations and procedural guarantees, as well as provisions for the assessment and mitigation of risks posed by AI systems. Also worth mentioning is the establishment of a monitoring mechanism in the form of a Conference of the Parties.
An important step towards AI that respects fundamental rights
The approval of an international convention focusing on AI is certainly a significant and necessary achievement. Indeed, the framework of shared fundamental principles and rights - which, however, build on the EU's catalogue of core values and principles - within which this technology is to be developed and used worldwide has been set in stone in international law.
The experience of Convention 108 on the Protection of Individuals with regard to Automatic Processing of Personal Data certainly speaks in favour of the importance of having legally binding international instruments in the context of human and anthropocentric governance of technologies.
Like any standard, this Council of Europe Framework Convention is not a perfect act and is not exempt from critical and improving remarks. Some have questioned some of the areas excluded from the scope of the treaty, some have rejected certain choices of approach and content. And of course, there are still many challenges and timeframes to be crossed before the real operational impact of this treaty can be seen.
However, the dimension of values, rights, and fundamental freedoms must always be the guiding star to be followed in the regulation and use of emerging technologies, and finally having a legally binding instrument to reaffirm and protect such a guarantee cannot but be an event to be enthusiastically welcomed.
Now we have to wait and see which and how many countries will sign the treaty and, above all, whether China, India, and the other Asian and Arab countries will join in, and, then, hope that Russia and gradually the emerging African powers will also join in. The road is long, but Europe has demonstrated once again, despite all its detractors, and with all the distinctions and problems we can list, that it is the only true human-centred lighthouse in our world.
⏰ That's all for us, see you insAIde, next time!
Rocco Panetta , Federico Sartore , Vincenzo Tiani, LL.M. , Davide Montanaro , Gabriele Franco