The Council of Europe Framework Convention on AI: what it requires and why it matters to lawyers

Alongside the European Artificial Intelligence Regulation there is another instrument, from the Council of Europe, with a different approach: an international treaty centred on human rights, democracy and the rule of law. It is the Council of Europe Framework Convention on artificial intelligence and human rights, democracy and the rule of law, number 225 in the Council of Europe Treaty Series.

According to its text, it opened for signature in Vilnius on 5 September 2024. And according to the Council of Europe itself, the European Union ratified it on 15 May 2026, during the 135th session of the Committee of Ministers, held in Chisinau.

What it aims to do

Article 1 sets its objective: to ensure that activities within the lifecycle of AI systems are fully compatible with human rights, democracy and the rule of law. Measures must be graduated according to the severity and probability of adverse impacts.

To whom it applies

Article 3 distinguishes:

• Public authorities, and private actors acting on their behalf: the convention applies to them directly.
• Private actors not covered by the previous case: each party must address their risks, applying the convention's principles or other appropriate measures, as it declares.
• Exclusions: national security activities, provided they respect international law and democratic institutions; research and development of systems not yet available, unless testing may affect rights; and national defence matters.

What is of particular interest to Justice

Article 5 obliges parties to protect democratic institutions and processes against uses of AI that undermine them. It expressly mentions the separation of powers, judicial independence and access to justice. For the legal profession, it is an international normative basis for demanding safeguards when AI is used in the Justice Administration, such as the AI use policy in the Justice Administration or the CGPJ instruction for judges.

The principles

The equality principle connects with what Spanish law already requires, as we explain when analysing Article 23 of Law 15/2022 on algorithms.

Remedies and safeguards

For lawyers, the most practical articles are 14 and 15:

• Effective and accessible remedies for human rights violations arising from AI.
• Documentation of relevant information about the systems, available to authorised bodies and, where appropriate, to affected persons, so they can challenge decisions.
• An effective complaint route before the competent authorities.
• Procedural safeguards when AI significantly affects rights.
• The aim that people know when they are interacting with an AI system and not a human being.

Risk management

Article 16 requires frameworks to identify, assess, prevent and mitigate risks, with monitoring, documentation and, where appropriate, testing before deployment. Parties must also consider whether moratoria, bans or other measures are needed for uses incompatible with human rights.

Entry into force

Under Article 30, the convention enters into force on the first day of the month following a period of three months after five signatories, of which at least three are Council of Europe member states, have expressed their consent to be bound.

The Treaty Office page with the status of signatures and ratifications could not be consulted at the time of writing, so we do not state here Spain's position or whether the convention is already in force. It is wise to check it on the Council of Europe website.

More analysis in Legal AI.