사회 가이드

유럽평의회 AI 프레임워크 협약

The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law is the first legally binding international treaty on AI.

4분 읽기
마지막 업데이트 2026년 9월 15일

읽기 시작 퀴즈 풀기

이 페이지에서4분 읽기

개요
심층 분석
전략적 영향
The Future of Council of Europe Framework Convention on AI
실제 구현
위험 및 가드레일
구현 로드맵
계속 탐색하세요
자주 묻는 질문

개요

Countries that ratify it must make sure AI systems used across their full lifecycle respect human rights, democratic processes and the rule of law. It matters because it sets shared legal duties that reach beyond Europe, and countries outside Europe can join it.

심층 분석

The Convention was negotiated by the Council of Europe's Committee on Artificial Intelligence (CAI). The Council of Europe's 46 member states took part, along with the European Union and observer countries such as the United States, Canada, Japan, Israel and Australia. The Committee of Ministers adopted the text on 17 May 2024, and it opened for signature in Vilnius on 5 September 2024. The first signatories included the United Kingdom, the United States, the EU, Norway, Israel, Iceland, Georgia, Moldova, Andorra and San Marino, and other states signed later. It comes into force after five ratifications, of which at least three must be Council of Europe member states. The treaty sets general principles rather than detailed technical rules. The principles are human dignity and individual autonomy, equality and non-discrimination, privacy and personal data protection, transparency and oversight, accountability and responsibility, reliability, and safe innovation. Parties must also provide remedies. That means documenting relevant information about AI systems, letting people challenge decisions, offering procedural safeguards, and telling people when they are dealing with an AI system. The scope is a compromise. The treaty applies fully to public authorities and to private actors acting on their behalf. For the rest of the private sector, each party can apply the treaty's obligations directly or take other appropriate measures. Activities to protect national security are excluded, as are matters of national defence and research and development on systems not yet made available for use. A common misconception is that the treaty is a copy of the EU AI Act. The AI Act is an EU regulation that sorts systems into risk tiers and sets product-safety style obligations. The Convention is an international human-rights instrument. The EU signed it and relies mainly on the AI Act and existing EU law to meet its duties.

기술적 통찰력

The core working duty is risk and impact management. Parties must adopt measures to identify, assess, prevent and mitigate risks to human rights, democracy and the rule of law across the AI lifecycle. These measures should consider how severe and how likely an impact is, and they should be repeated as the system changes. To support this, the CAI adopted HUDERIA in late 2024. HUDERIA is a non-binding methodology for assessing the risks and impacts of AI systems on human rights, democracy and the rule of law. For practitioners, the treaty mostly arrives through national law: procurement rules, algorithmic impact assessments, logging and documentation duties, and complaint routes. Parties must also consider whether to ban or pause certain uses they judge incompatible with human rights. A Conference of the Parties monitors implementation and supports cooperation, and parties report to it periodically. Because the text is technology-neutral, the concrete obligations depend on how each country implements it.

전략적 영향

위험과 안전

치명적인 AI 피해와 일상적인 AI 피해는 누가 위험을 이해하고 누가 조치를 취할 수 있는지에 따라 달라집니다.

더 명확한 결정들

공공 및 전문 지식은 강력한 안전 정책이 정치적으로 가능한지 여부를 결정합니다.

과장된 과장을 뚫고 나가기

명확한 설명은 과대광고, 연구실 홍보, 모호한 윤리 연극에 의한 포착을 줄입니다.

The Future of Council of Europe Framework Convention on AI

How much the Convention matters will depend on ratifications and on how strictly countries treat the private sector. Countries that choose "other appropriate measures" may implement it quite differently from countries that apply it directly to companies. The Conference of the Parties and tools like HUDERIA could slowly build shared expectations, much as earlier Council of Europe treaties did for data protection. Watch for which non-European countries ratify it, how national courts use it, and how it interacts with the EU AI Act and national AI laws. Its reach is broad, but enforcement relies on national systems rather than an international regulator.

실제 구현

A government agency that uses an AI tool to flag benefit fraud would need risk and impact assessments, documentation, and a way for affected people to challenge decisions that significantly affect their rights.

A citizen who talks with a public-service chatbot should be told that they are interacting with an AI system rather than a human.

A ratifying country may decide that private companies using AI in hiring have to meet the treaty's principles directly, or it may take other appropriate measures to meet the same goals.

An election authority worried about AI-driven disinformation can point to the treaty's duty to protect the integrity of democratic processes when it designs safeguards.

위험 및 가드레일

실존적 위험을 공상과학처럼 다루면서 능력을 합성합니다.
높은 자율성 하에서 정렬과 표면 제품 안전성을 혼동합니다.
영어가 아니거나 전문가가 아닌 청중에게는 품질이 낮은 소스만 남겨 둡니다.

구현 로드맵

제품 손상, 오용, 통제력 상실/잘못 정렬 위험을 분리합니다.
일정과 심각도에 대한 귀하의 견해를 바꿀 수 있는 증거가 무엇인지 물어보십시오.
마케팅 주장보다 기본 소스와 구체적인 평가를 선호하세요.
인식뿐만 아니라 경력, 정책, 자금 조달 또는 기술 등 하나의 행동 경로를 식별하십시오.

계속 탐색하세요

Free newsletter

Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.

One email each weekday. Unsubscribe in one click. We never sell or share your address.

Test yourself

Take the Council of Europe Framework Convention on AI quiz

Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.

퀴즈 시작

Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation

자주 묻는 질문

What is Council of Europe Framework Convention on AI?

The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law is the first legally binding international treaty on AI. Countries that ratify it must make sure AI systems used across their full lifecycle respect human rights, democratic processes and the rule of law. It matters because it sets shared legal duties that reach beyond Europe, and countries outside Europe can join it.

What makes the Council of Europe Framework Convention on AI historically notable?

The Convention is the first international AI treaty that binds the states that ratify it. Voluntary codes and EU regulations are different kinds of instruments.

Where did the Convention open for signature in September 2024?

It opened for signature in Vilnius, Lithuania, on 5 September 2024, during a Council of Europe ministerial conference.

How does the treaty handle private companies that are not acting for public authorities?

This was a negotiated compromise. The public sector is fully covered, and each state chooses how to address other private-sector AI use.

Which activities are excluded from the treaty's scope?

Activities to protect national security and national defence matters are excluded. So is R&D on systems not yet made available for use.

What is HUDERIA?

HUDERIA was adopted by the CAI to help assess AI impacts on human rights, democracy and the rule of law. It is guidance, not a binding part of the treaty.

계속 학습하세요