Our trust design
Every ethical decision behind SpeakForward is the product itself, not a compliance burden. These decisions were finalized before development began — they are not traded away under deadline pressure.
The platform is open to any 18–25 year-old leader who cannot speak or record naturally due to a health condition: vocal cord injury, severe anxiety, ALS, stroke recovery, chronic fatigue syndrome, and more. We do not require medical proof — medical proof is itself a privilege, and it gates out exactly the people who most need to be heard. Eligibility is confirmed by a short public statement; the judgment of trust is delegated to the community and donors.
The platform asks only for an email and an age self-declaration. No KYC, no phone verification, no OAuth. "This platform does not verify identities. Please judge for yourself." is displayed site-wide, permanently. We chose to make fraud risk visible rather than to filter out real people with screening walls.
Every voice is generated by ElevenLabs with the native AI-disclosure flag embedded in the request, and every page carries a visible label: "This content is AI-generated, on behalf of a creator with a speech disability". We never pretend to be a human voice — disclosure is respect for the listener.
The health statement a creator submits is displayed publicly on the project page. That is not a privacy violation — it is the foundation of community trust. Donors see a real person with a name and a situation, not an anonymous ask.
Donations are purely peer-to-peer: SOL moves from the donor's wallet directly to the creator's wallet, signed by the donor's own Phantom. The platform never holds, routes, or custodies funds. Donors without a wallet can use a 'pledge' — stored only in their own browser, clearly labeled as simulated.
Every content anchor (the creator's Memo transaction carrying a SHA-256 content hash) and every donation signature is public and independently verifiable on the Solana Devnet explorer. Trust doesn't rest on the platform's promises — it rests on publicly auditable on-chain facts. No custom Solana programs are deployed, keeping the attack surface and audit burden at zero.