Likely AI-generated. Pangram flags 100% of this text as AI-written.
1
Independent Multi-Model AI Accountability Research by the EM Foundation
ProposalGrant
Closes September 14th, 2026$0raised
$75,000minimum funding
$1,500,000funding goal
Offer to donate
37 daysleft to contribute
You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
Project summary
EM Foundation for AI Research is a Texas 501(c)(3) (EIN 42-3000086) doing independent, adversarial research on AI accountability, checking one AI system's reasoning against another's instead of a lab checking its own. Since incorporating in May 2026, working solo and founder-funded, I've built a working multi-model deliberation platform, published several adversarially reviewed Research Notes, and completed a Working Paper that passed a seven-role hostile review. I'm seeking funding to move this from part-time and founder-funded to durable.
What are this project's goals? How will you achieve them?
Goal: make independent, cross-model AI accountability research sustainable past a single founder working part-time. Path: reliable always-on hosting for the deliberation engine and research tools, funded compute time on the quantum backends already integrated (IBM Quantum, AWS Braket), and bringing in outside reviewers to scale the adversarial-review process that gives the published work its credibility.
How will this funding be used?
[Fill in your real numbers here, this is the section that needs your actual figures, not mine. Rough categories to cost out: hosting/infra migration, quantum compute time, reviewer/collaborator capacity. State a total ask and how it splits.]
Who is on your team? What's your track record on similar projects?
Solo founder currently: Desmond Iwuagwu, President. Track record: multiple published Research Notes that went through structured adversarial review, including one AI system independently reviewing deliberation research produced by others; a completed Working Paper that survived seven-role hostile review with every major criticism addressed; three provisional patents filed pro se with honest self-assessed patentability; working infrastructure already live and integrated with real backends, not a proposal-stage idea.
What are the most likely causes and outcomes if this project fails?
Most likely failure mode isn't the research being wrong, it's scale: staying a part-time, one-person effort indefinitely because infrastructure and compute costs outpace what a small grant covers, or the published research not finding an audience/reviewer base beyond me. Funding this reduces the first risk directly; the second is a real, ongoing risk regardless of funding.
How much money have you raised in the last 12 months, and from where?
$0 external, fully founder-funded to date.