IAIA Blueprint Summary Proposal
This paper proposes a practical blueprint for an International Artificial Intelligence Agency (IAIA) focused on compute governance of frontier AI training runs. It sets out six operational questions, a three-tiered guardrail, a grand-bargain incentive structure and graduated responses, using the concentrated chip supply chain and unresolved epistemic risk as the basis for oversight.
Authors
- Karen Laing
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-14
- DOI
- https://doi.org/10.5281/zenodo.24223107
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- preprint