Methodology

What is an AI Governance Venue (AGV)?

An AGV is any organized forum, body, initiative, coalition, or recurring convening whose mandate includes — as a primary or substantial secondary purpose — the governance, oversight, coordination, or norm-setting of AI technologies, their development, deployment, or societal impact.

The term venue deliberately covers continuous organizations (e.g., OECD.AI), recurring fora (e.g., AI Action Summits), one-off convenings that produced durable outputs (e.g., Asilomar 2017, Bletchley Park 2023), and workstreams within larger organizations that have a distinct AI mandate.

Inclusion and exclusion

Excluded from the dataset:

• Purely technical AI conferences without any policy or governance track.
• National-only entities with no international participation or extraterritorial reach (with the exception of national AISIs, included for their explicit role in the international AISI network).
• Corporate internal AI ethics teams without external coordination function.

Classification dimensions

Each venue is classified across six orthogonal dimensions defined in the AGVO ontology v0.1:

1. Entity type — twelve values from igo_initiative to one_off_summit.
2. Governance modality — what the venue produces: declarations, binding instruments, technical standards, evaluations, capacity-building, research, dialogue, regulation.
3. Topic focus — what aspect of AI: general, safety/frontier, ethics/rights, privacy/data, generative AI, agentic AI, or domain-specific.
4. Geographic scope — global, transregional, regional, plurilateral, or bilateral-plus.
5. Lead actor type — who drives it: IGO secretariat, state government, industry, academia, civil society, or hybrid multistakeholder.
6. Legal character — soft law, hard law, mixed, or n/a.

A separate lifecycle model tracks state transitions: announced → active → optionally dormant, absorbed, terminated, or succeeded.

Data sources

Current sources for the seed sample:

• Direct venue homepages and founding documents.
• OECD.AI Policy Navigator.
• UNESCO Global AI Ethics and Governance Observatory.
• IAPP Global AI Law and Policy Tracker.
• Berkman Klein "Principled AI" mapping (historical).

Future automated ingestion is planned via dedicated fetchers per source.

Update cadence

Monthly. The pipeline runs on the first of each month, fetches changes from upstream sources, and opens a Pull Request listing candidate additions and updates. Each change is human-reviewed before being merged into the canonical dataset.

Limitations

This is a v0.1 pilot release with a hand-curated seed sample of 30 venues. The selection is not statistically representative; it is a stress test of the ontology. Coverage gaps that will be addressed in future releases include:

• Industry consortia formed in 2024–2025 (only the most prominent are included here).
• Regional bodies in Africa, Latin America, ASEAN beyond their largest convenings.
• Standards bodies beyond ISO/IEC SC 42 (IEEE working groups, ITU AI focus groups).
• Many academic conference policy tracks.

Cite this release

Source code

Repository: https://github.com/s00048ri/agv-tracker. See CITATION.cff at the repo root for the GitHub-native citation widget, and .zenodo.json for the Zenodo deposit metadata.