The honest critique of how giving works — and what we built instead.
Every platform in this space solves part of the problem. None solve all of it. Here's what we learned — and what we built.
Funds move before anything is verified.
On every major giving platform, donations transfer before independent verification occurs, if verification occurs at all. Progress reports are self-submitted by the organizations receiving the money. That’s a structural conflict of interest.
Escrow + Independent AI Verification
Every donation enters escrow. Not a cent moves until our AI independently reviews the evidence — photos, GPS data, community voice — and scores the milestone. The implementing org has no role in their own verification.
Communities are the subject, not the author.
Charity: water’s films are beautiful. GlobalGiving’s partner updates are well-intentioned. But in every case, a communications team decides what gets told, how it’s framed, and who’s photographed. Communities are the content — not the storytellers.
Community Voice as Infrastructure
Community members record voice notes in their language. Our AI transcribes, translates, and publishes them as the donor impact story. The community approves the final version. They own their data. The organization does not.
Platform fees punish donors and partners alike.
Traditional NGOs take 20–40% overhead before any program dollar moves. Crowdfunding platforms charge 8–15% at the moment of donation — not when impact is delivered. Some emerging platforms add their fee on top of each milestone payment. Our fee is detached from whether anything worked.
5–8% from Escrow, Only on Verification
We take our fee from inside the escrow, only when a milestone is verified. If a project fails, no Firsthand fee is charged. The fee structure aligns our incentives with actual outcomes — not with the volume of donations processed.
AI is being applied to fundraising copy, not accountability.
Most platforms using AI are using it to generate fundraising emails, optimize donation pages, or write impact summaries. The accountability chain — where trust actually breaks — remains untouched.
AI at Every Stage of the Accountability Chain
Firsthand uses AI for: Gate 1 application screening (5-dimension scoring), milestone evidence verification (photo, GPS, voice cross-reference), community voice transcription and translation, and donor-project matching. AI is the verification layer — not the marketing layer.