Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law
Council of Europe · Law · Council of Europe
High impact. A reviewer confirms each change before it is listed here.
Our summary
This international treaty sets binding principles and requirements for member states to implement regarding AI systems throughout their lifecycle. It applies to government authorities and, to varying degrees depending on national implementation, to private actors, while excluding national defence and some research activities. Parties must adopt measures ensuring AI activities comply with existing obligations on fundamental freedoms, democratic governance, non-discrimination, accountability and data protection.
Written by the Index in our own words. It is not legal advice; the official text is the authority.
Inside this policy
Every section of this policy in order. Our notes summarise each one in plain language and do not quote the official text, which is the authority.
Drafted by the Index An automated summary of the official text, not yet reviewed by a person.Reviewed Checked by a person against the official text.
44 sections
Jump to a chapter
- Chapter I General provisions
- Chapter II General obligations
- Chapter III Principles related to activities within the lifecycle of artificial intelligence systems
- Chapter IV Remedies
- Chapter V Assessment and mitigation of risks and adverse impacts
- Chapter VI Implementation of the convention
- Chapter VII Follow-up mechanism and cooperation
- Chapter VIII Final clauses
Chapter I General provisions
3 sections
Chapter I introduces the framework covering the definition and scope of artificial intelligence systems, establishing foundational concepts for the Convention on aligning AI with human rights, democracy and the rule of law.
Article 1 Object and purpose
Parties must enact or maintain measures to align AI system activities throughout their lifecycle with fundamental rights, democratic principles and rule of law. Measures must be scaled and adapted based on the severity and probability of potential harm. This covers specific or broad measures regardless of the technology involved.
Article 2 Definition of artificial intelligence systems
An AI system is a machine-based tool designed for explicit or implicit goals that learns from input data to create outputs like forecasts, generated material, suggestions or determinations affecting physical or digital contexts. Such systems differ in their degree of independence and adaptability following deployment.
Article 3 Scope
The Convention applies to AI lifecycle activities by public authorities and private actors on their behalf, and to private actors' AI risks not covered by public authorities, unless related to national security, defence or research not yet available for use without testing that could interfere with rights and democracy.
Chapter II General obligations
2 sections
Chapter II sets out Parties' core obligations to protect human rights and safeguard the integrity of democratic institutions and rule of law from AI systems' impacts.
Article 4 Protection of human rights
Parties must ensure that AI lifecycle activities comply with international and domestic human rights law obligations.
Article 5 Integrity of democratic processes and respect for the rule of law
Parties must adopt measures so AI systems do not undermine democratic institutions, separation of powers or judicial independence, and must protect fair access to public debate and individuals' freedom to form opinions.
Chapter III Principles related to activities within the lifecycle of artificial intelligence systems
8 sections
Chapter III establishes general common principles Parties must implement for AI systems covering human dignity, transparency, accountability, equality, privacy, reliability and safe innovation, applied appropriately to their legal systems.
Article 6 General approach
Parties shall implement the Chapter III principles regarding AI systems in a manner suited to their domestic legal framework and other Convention obligations.
Article 7 Human dignity and individual autonomy
Parties must adopt or maintain measures protecting human dignity and individual autonomy throughout AI lifecycle activities.
Article 8 Transparency and oversight
Parties must adopt or maintain measures requiring adequate transparency and oversight tailored to specific contexts and risks throughout the lifecycle of AI systems, including identification of AI-generated content.
Article 9 Accountability and responsibility
Parties must adopt or maintain measures ensuring accountability and responsibility for harms to human rights, democracy, and rule of law caused by activities within the AI system lifecycle.
Article 10 Equality and non-discrimination
Parties must adopt or maintain measures ensuring AI system lifecycle activities respect equality including gender equality and prohibit discrimination under applicable law, and overcome inequalities to achieve fair and equitable outcomes.
Article 11 Privacy and personal data protection
Parties must adopt or maintain measures protecting individuals' privacy rights and personal data in AI system lifecycle activities through applicable laws and standards, and ensuring effective safeguards for individuals.
Article 12 Reliability
Parties shall as appropriate take measures promoting AI system reliability and user trust, which may include requirements for adequate quality and security throughout the AI system lifecycle.
Article 13 Safe innovation
Parties are called upon to as appropriate enable controlled environments supervised by competent authorities for developing, experimenting and testing AI systems, balancing innovation with avoiding harms to human rights, democracy and rule of law.
Chapter IV Remedies
2 sections
This chapter covers remedies and procedural safeguards for violations resulting from AI system lifecycle activities, requiring accessible and effective remedies, information documentation and complaint mechanisms.
Article 14 Remedies
Parties must ensure accessible and effective remedies for human rights violations from AI system activities. They must document and provide relevant information to authorized bodies and affected persons sufficient for contesting AI-based decisions, and enable complaint lodging with competent authorities.
Article 15 Procedural safeguards
Parties must provide effective procedural guarantees and safeguards to persons significantly impacted by AI systems on human rights enjoyment. Parties should seek to notify persons they interact with AI systems rather than humans.
Chapter V Assessment and mitigation of risks and adverse impacts
1 section
This chapter covers a risk and impact management framework for assessing and mitigating risks and adverse impacts from AI systems.
Article 16 Risk and impact management framework
Parties must establish measures to identify, assess, prevent and mitigate risks posed by AI systems to human rights, democracy and rule of law, tailored to context, severity and probability of impacts. Measures must apply throughout the system's lifecycle, include stakeholder perspectives, monitoring, documentation, and pre-deployment testing where appropriate. Parties must address adverse impacts and may impose bans on incompatible uses.
Chapter VI Implementation of the convention
6 sections
This chapter covers Articles 17 to 22, establishing principles for implementing the convention: non-discrimination, protecting persons with disabilities and children, public consultation, digital literacy, safeguarding existing rights, and allowing wider protections.
Article 17 Non-discrimination
Parties must implement this convention without discrimination on any ground, consistent with their international human rights obligations.
Article 18 Rights of persons with disabilities and of children
Parties must account for the specific needs and vulnerabilities of persons with disabilities and children when implementing the convention, according to their domestic law and international obligations.
Article 19 Public consultation
Each Party shall encourage that key matters concerning AI systems be examined through public dialogue and broad stakeholder engagement, considering their implications across societal, business, judicial, moral, environmental and other important dimensions.
Article 20 Digital literacy and skills
Parties shall encourage and promote digital literacy and skills across the population, including specialized expertise for persons responsible for identifying, assessing, preventing and mitigating risks from AI systems.
Article 21 Safeguard for existing human rights
The convention does not limit, derogate from or affect human rights or legal rights and obligations guaranteed under a Party's laws or other international agreements to which it is party.
Article 22 Wider protection
Parties may grant more extensive protections than required by this convention.
Chapter VII Follow-up mechanism and cooperation
4 sections
This chapter covers Articles 23 to 26, establishing the mechanism for ongoing implementation: the Conference of the Parties, reporting obligations, international cooperation, and oversight mechanisms.
Article 23 Conference of the Parties
The Conference comprises Party representatives and meets periodically to facilitate implementation, consider amendments, exchange information, settle disputes, and cooperate with stakeholders. The Secretary-General convenes it when necessary or on request by a majority of Parties or the Committee of Ministers. Parties shall adopt rules within 12 months.
Article 24 Reporting obligation
Each Party must submit a report within two years of accession and periodically thereafter, detailing its actions to implement the key obligations in Article 3. The Conference of the Parties will set the reporting format and process.
Article 25 International cooperation
Parties must cooperate to achieve the convention's purpose and are encouraged to assist non-parties in aligning with it. Parties must exchange information on AI's effects on human rights, democracy and rule of law, including risks and private sector impacts, and strengthen cooperation to prevent and mitigate adverse effects.
Article 26 Effective oversight mechanisms
Each party must establish or designate one or more independent mechanisms to oversee compliance with the convention's obligations, equipped with necessary powers, expertise and resources. Where multiple mechanisms exist, parties must facilitate cooperation among them and between them and existing human rights structures.
Chapter VIII Final clauses
10 sections
This chapter covers final procedural and administrative provisions governing the convention's effects, amendments, dispute resolution, signature and entry into force, accession, territorial application, federal application, reservations, denunciation and notification procedures.
Article 27 Effects of the Convention
Parties may apply existing bilateral agreements or treaties on the convention's subject matter if consistent with its object and purpose. EU member parties apply EU rules among themselves without prejudice to the convention's full application with other parties.
Article 28 Amendments
Any Party, the Committee of Ministers or the Conference may suggest amendments, communicated by the Secretary-General. The Conference advises the Committee on proposals; the Committee may approve amendments and forwards them to Parties for ratification. Approved amendments take effect when all Parties accept them, on the thirtieth day after final notification.
Article 29 Dispute settlement
When Parties disagree about the Convention's meaning or application, they shall resolve disputes through negotiation or other peaceful methods they select, such as through the Conference of the Parties.
Article 30 Signature and entry into force
Council of Europe members, non-member states that contributed to drafting, and the European Union may sign the Convention. It requires ratification and enters into force three months after five signatories, including at least three Council members, deposit their ratification. For later signatories, it enters into force three months after their individual ratification.
Article 31 Accession
After entry into force, the Committee of Ministers may, with unanimous party consent, invite non-member states that did not participate in drafting to accede. The convention enters into force for acceding states three months after depositing their accession instrument.
Article 32 Territorial application
Parties may specify the territories to which the convention applies at signature, ratification or accession, and may extend or withdraw application to additional territories with three months' notice to the Secretary-General.
Article 33 Federal clause
Federal states may adjust their obligations to align with their constitutional arrangements governing central and regional authority, though the Convention applies to the central government. Federal authorities shall inform and encourage constituent states to implement the Convention where they have legislative responsibility.
Article 34 Reservations
States may make a written reservation regarding the federal clause when signing or ratifying the Convention by notifying the Council of Europe Secretary-General. No other reservations are permitted.
Article 35 Denunciation
A Party may withdraw by written notice to the Secretary-General at any time. Withdrawal takes effect three months after the Secretary-General receives the notice.
Article 36 Notification
The Council of Europe Secretary-General must notify relevant states, the European Union and other parties of signatures, ratifications, entry into force, amendments, declarations, reservations, withdrawals, denunciations and other acts relating to the Convention.
Controls
What this policy requires, forbids, grants or penalises, one control at a time and in our own words. Each control links to the official text, which is the authority. This is not legal advice.
Drafted by the Index Drafted automatically from the official text, not yet reviewed by a person.Reviewed Checked by a person against the official text.
Themes covered 0 of 35
Choose a theme to see only its controls. About the themes
Risk and scope 0 of 4
- Risk classification, not covered
- Prohibited practices, not covered
- Risk management system, not covered
- Impact assessment, not covered
Data and privacy 0 of 4
- Data governance and quality, not covered
- Data subject rights and consent, not covered
- Cross-border data transfer, not covered
- Copyright and training data, not covered
Transparency and records 0 of 4
- Transparency and disclosure to users, not covered
- Labelling of synthetic content, not covered
- Technical documentation, not covered
- Record keeping and logging, not covered
Technical safeguards 0 of 4
- Human oversight, not covered
- Accuracy, robustness and cybersecurity, not covered
- Bias and non-discrimination, not covered
- Testing and evaluation, not covered
Assurance and lifecycle 0 of 4
- Conformity assessment and certification, not covered
- Registration and notification, not covered
- Incident and breach reporting, not covered
- Post-market monitoring and corrective action, not covered
Rights and protected groups 0 of 4
- Automated decision-making rights, not covered
- Children and vulnerable groups, not covered
- Biometric identification and surveillance, not covered
- Whistleblowing, complaints and redress, not covered
Roles and organisation 0 of 6
- Supply chain and provider-deployer duties, not covered
- General purpose and foundation models, not covered
- Accountability and governance roles, not covered
- Training and AI literacy, not covered
- Public sector use and procurement, not covered
- Innovation support and sandboxes, not covered
Enforcement and sectors 0 of 5
- Enforcement, penalties and regulators, not covered
- Sector rules: health, not covered
- Sector rules: finance and insurance, not covered
- Sector rules: employment, not covered
- Sector rules: education, not covered
The full list of controls could not be loaded, so only the first 0 are shown.
Change history
No changes recorded since we began monitoring this policy.