OECD Working Party and Network of Experts on AI

The Working Party on Artificial Intelligence Governance oversees the OECD's work on AI policy. The OECD.AI Network of Experts provides policy, technical and business expert input to inform OECD analysis and recommendations.

Working Group on Classification & Risk

Different types of AI systems raise very different policy opportunities and challenges. The Working Party on AI Governance has an expert group that works on the classification of AI systems. To date, it has developed a user-friendly framework to classify AI systems.

The OECD AI Systems Classification Framework provides a structure for assessing and classifying AI systems according to their impact on public policy in areas covered by the OECD AI Principles: economic and social benefits; human rights, privacy and fairness; safety, security, and risk assessment; transparency; accountability; research; data, compute and technologies; labour and skills; and international cooperation.

The Framework builds on the conceptual view of a generic AI system established in previous OECD work. It identifies policy considerations associated with different AI systems’ attributes including the:

  1. System’s socio-economic context including its sector, application, and whether it constitutes a critical activity;
  2. Data/input of the AI system;
  3. AI model/technologies in use; and
  4. Task and action of the AI system.

Co-chairs and members

The classification expert group is co-chaired by:

Marko Grobelnik (AI Researcher & Digital Champion, AI Lab, Slovenia Jozef Stefan Institute);

Dewey Murdick (Director of Data Science, Center for Security and Emerging Technology (CSET), School of Foreign Service, Georgetown University); and

Sebastian Hallensleben Head of Digitalisation and AI – VDE Association for Electrical, Electronic & Information Technologies and the chair of CEN-CENELEC JTC 21

The group meets virtually every 3 to 4 weeks.

See Working Group participants

Blog post

Documents

Business

Putting the OECD AI Principles into practice: progress and future perspectives – Event summary

Session summary of this high-level panel discussion from the OECD MCM 2021 More than two years after the adoption of the OECD AI Principles, this discussion will highlight how countries are implementing policies for trustworthy, human-centric AI. The session will focus on: - Lessons learned to date implementing AI policies to steer the transformation of our economies and societies for inclusive growth and addressing global challenges. - Risk-based approaches to AI governance and soft and hard law, including the proposed EU AI Act that is triggering discussions around the world. - How to ensure that our policies and legal frameworks for AI are interoperable globally. And the priorities for international cooperation on AI moving forward.June 17, 2022

Intergovernmental

The technological readiness level (TRL) of 66 initiatives grouped based on the clustering framework described in Responsible AI in Pandemic Response

The technological readiness level (TRL) of 66 initiatives. Initiatives are grouped based on the clustering framework described in Responsible AI in Pandemic Response (The Future Society, 2020). Visualization by Bruno Kunzler, TFS Affiliate.April 6, 2022

OECD Framework for Classifying AI Systems: two page overview

The framework classifies AI systems along five dimensions: People & Planet, Economic Context, Data & Input, AI Model and Task & Output. Each one has its own properties and attributes that help assess policy considerations of particular AI systems. Stakeholders are involved in or affected by AI systems, while AI actors play active roles according to each dimension and throughout an AI system’s lifecycle. The framework allows users to zoom in on specific risks that are typical of AI, such as bias, explainability and robustness yet it is generic in nature. It facilitates nuanced and precise policy debate. The framework can also help develop policies and regulations since AI system characteristics influence the technical and procedural measures they need for implementation. The framework promotes a common understanding of AI by identifying the features of AI systems that matter most. Both in and out of government, the framework can inform registries or inventories by guiding work to describe systems and the basic characteristics of algorithms or automated decision systems. February 18, 2022

Business

The OECD Framework for the Classification of AI systems

Key dimensions structure AI system characteristics and interactions The framework classifies AI systems and applications along the following dimensions: People & Planet, Economic Context, Data & Input, AI Model and Task & Output. Each one has its own properties and attributes or sub-dimensions relevant to assessing policy considerations of particular AI systems. February 17, 2022

Sign up for OECD artificial intelligence newsletter