The development of a legal framework for the use of AI is still at an early stage. Moving forward, it is necessary to take into account both the inherent features of the technology and the rights that come under pressure by our use of it. The approach of the European Court of Human Rights (ECtHR) will have to be on a case-by-case basis, building on the Convention’s transversal values, applying existing jurisprudence as a stepping stone and making wise use of the “living instrument doctrine”.

We have just started

The legal landscape surrounding AI is still in its initial stages, both in terms of applicable regulations and case-law. It was only in 2024 that the Council of Europe and the European Union adopted the first legally binding instruments aimed at providing a legal framework for the States to regulate the activities within the lifecycle of AI systems and their impact on democracy, the rule of law and the protection of human rights.

Consequently, in addition to the galloping technical development and the rapidly expanding use – and abuse – of it, any analysis of AI and its implications must be approached with an understanding of the evolving legal regime and the potential for significant developments on that front also.

From the point of view of Convention case-law itself, none of the Court’s judgments to date, which could be considered “key” as regard digital/technological advancement, concern artificial intelligence as such.

Indeed, by the nature of the system, Convention cases will not preface the development of AI. And they will as such emerge only after the application of national AI rules, regulations and protections, and of the more specific international regulations, as just mentioned, such as under EU law. In this respect, the proper implementation of the Convention requires that the Member States take the lead, going first in securing that AI is not applied to the detriment of Human Rights.

This is also the only viable approach, considering the principles of subsidiary and shared responsibility as reinforced by Protocol 15 to the Convention and in the Court’s more recent case law.

The inherent risks with AI

Turning to the main risk posed by AI vis-à-vis the enjoyment of Human Rights, I believe we need to start with recognising that the technology has certain inherent features worthy of attention, notably the following four.

1. It is a “black box”. Very few, if anyone, can fully comprehend how it operates – the algorithms and their ways of processing information remain in the dark.
2. It cannot be trusted. AI has fundamentally no real sense of truth and not truth. The problem of hallucination is severe. What a human would call a lie or pure nonsense is presented very convincingly as trustworthy information.
3. AI technology is, by its very nature, very adaptive and effective. It can be extremely intrusive. And the potential of unwise use is virtually unlimited. Not to speak of abuse.
4. The costs of developing and maintaining AI technology are huge. The technology is likely to be controlled by big companies and investors, whose activities may escape effective public oversight.

The rights at play

Turning more concretely to the rights at play, the list could probably be virtually endless. I will mention here in particular the following:

• The erosion of the right to privacy;
• Structural harms affecting human dignity (for example, by way of surveillance technologies eroding human autonomy, self-governance and self-determination);
• The aggravation of existing discrimination and inequalities;
• The challenges to freedom of expression and of information (AI-driven moderation, algorithms with addictive designs/echo chambers);
• The threats to transparency and accountability;
• Attacks on the integrity of the electoral process, by interference from foreign hostile powers and the manipulations of elections;
• The blurring of the very notions of truth and untruth, the rejection of science-based knowledge, turning any issue into a question of opinion.

The way ahead

As we proceed, the Court will have to develop and clarify the Convention principles faced with, and adapted to, AI on a case-by-case basis. That implies that there will be no grand Master Plan, according to which the Court will design its jurisprudence. Law is, on the contrary, created cautiously, step by step, within the boundaries of the Court’s competences and jurisdiction under the Convention and in accordance with recognised principles of adjudication and legal method.

This task needs to be carried out at all levels of decision-making, notably at the level of the Grand Chamber – be it via individual complaints or via requests for advisory opinions under Protocol 16 to the Convention.

On this venture into uncharted legal waters, the Court will – I assume – have to seek guidance in well-known general tools and methods, such as the following:

First: Understanding the facts, the technology and the challenges and benefits affiliated with it, notably those relevant to the protection of Human Rights.

Second: The Court must be clear that any emerging new case law on AI must be compatible with the continued effective protection of the Convention’s transversal constitutive values, notably on the Rule of law, on Democracy, and on Human Dignity.

Third: In clarifying the functioning of the Convention in respect of AI, the Court will draw as much as possible from already well-established case law in respect of how to deal with the challenges of new technologies, for example such as on internet safety and the right to be forgotten, on responsible journalism in the online sphere, on data transfer, the protection of journalistic sources, on the right to free elections, on fair trial and on the prohibition of discrimination. From this huge body of case law viewed on an aggregated level, certain tendencies, values, mechanisms or principles may most likely be extracted.

Fourth: The Court will apply the living instrument doctrine to make sure that the Convention right in question remains effective in practice and that the interpretation and application of the Convention are adapted to the present-day conditions.

Indeed, without this approach to Convention interpretation, the Convention would already have been completely outdated and without any real bearing for such drastic leaps in societal development as AI. In this respect, I would expect that the Court will – when defining the direction of speed of any jurisprudential development – take into account the possible emerging consensus in domestic law of the Member States of the Council of Europe, within the European Union, under other international Human Rights instruments and in international law more broadly.