Describe (a) what you plan to do, (b) why you believe it is valuable, and (c) the main downside risks and how you will mitigate them. If you have a particular timeline in mind for the project, please mention that. We recognize this may not be enough space to fully explain your proposal; this is just to allow initial screening, and we may ask follow-up questions to applicants as needed.

What we plan to do

We will run a series of in-person convenings, each targeted at one specific mechanism of AI-driven power concentration (e.g. regulatory capture, surveillance infrastructure, compute/model-access control). Every event integrates all three asset types this priority area names: training (facilitated, practice-based sessions drawing on established social-innovation methods), the event itself (a working session building a real local case study, not an abstract briefing), and a public resource (a rich, multi-layered case study produced via our “Publishing Stack” - infrastructure that captures and structures on-the-ground organising data live, rather than as a post-hoc write-up). Because production runs on lightweight, local-first tools, making each case study is itself a small working demonstration of decentralized information governance - giving readers a second, meta layer to learn from alongside the substantive content. Case studies are self-guided, multi-perspective learning environments; facilitation templates are published openly for other organizations to run themselves.

Why this is valuable

This area needs both breadth (several under-examined mechanisms) and depth (people who can actually contribute, not just read about it). Structuring the series as narrowly-targeted individual events - each still producing training, convening, and resource - covers breadth across the series without diluting any single session. Our infrastructure produces case studies with a density and provenance most content programs can’t match, since the data comes directly from live organising rather than a summary written afterward - and the production method doubles as substantive material on AI/data governance in practice.

Risks and mitigation

Track record gap: our convening experience is adjacent to AI governance (governance and power, infosec, decentralized organizing), not squarely inside it; mitigated by co-facilitation and review from established AI governance researchers on the pilot event before wider publication. Attendance without follow-through: mitigated by a funded fellowship track with to retain contributors.

Timeline: pilot event within 3-4 months of funding; 3 further targeted events and published case studies over the following 6-9 months.