Authenticated messaging

Check the photo,not the sender.

Chainmail is an encrypted messenger that can stamp the photos you send with a fingerprint recorded on Solana (devnet today). If someone edits the file later, anyone can see it — in their own browser.

Messages are end-to-end encryptedPhoto stamps are checked by open-source code
IMG_2048.JPGExample
IH1 · STAMPED
IntactFingerprint matches the Solana record
Recorded2026-10-14 09:42
Hash8f3a…9d8c
Checkedin browser
Proves: unchanged since stamped. Does not prove: who took it, or that a camera made it.
The problem

Seeing is no longer believing.

AI can make a convincing photo in seconds, and most people cannot tell. When a picture arrives in a chat, the only question that matters is whether it is the picture that was taken.

53%

of US adults are not confident they can spot AI-made content

Pew Research Center, Sep 2025
76%

say it matters to tell AI content from human content

Pew Research Center, Sep 2025
0.1%

of 2,000 people identified every real and fake item in a test

iProov, Feb 2025 (vendor study)
Why current fixes fall short

Guesses, stripped data, and trust.

01

Detectors guess

AI detectors give a probability, and they fall behind each new model.

02

Metadata is easy to strip

Most apps re-encode photos and drop provenance data on the way.

03

Badges ask for trust

A platform's checkmark is a promise from the platform, not something you can test.

The solution

A stamp you can check without trusting us.

Chainmail seals three hashes into each stamped photo and records them on Solana (devnet today). Our open-source verifier re-derives the result from the file itself, so the answer never depends on our servers.

Edits show

Change one byte after stamping and the check fails.

Private by default

Messages in one-to-one and group chats are end-to-end encrypted.

Checked locally

The web checker runs in your browser. The photo does not upload.

Scope stated

Every result says what it proves and what it does not.

Features

What ships today, and what is next.

Live

Stamped photos

Turn on stamping and each JPEG you send gets an IH1 fingerprint with a memo on Solana (devnet today). Recipients see “Hashed on-chain”.

MayaMessages end-to-end encrypted
Today 09:41
Is this the house? The listing photos look different.
Hashed on-chain
Took it just now. Check the stamp.
Live

Encrypted messages

X3DH and Double Ratchet with XChaCha20-Poly1305. Groups use sender keys.

Your phone

identity key
ed25519 · on device
X3DH · Double Ratchetxchacha20-poly1305
3f9a…c1e0 7b2d…
server sees ciphertext

Their phone

identity key
ed25519 · on device
Live

Open verifier

@chainmail/verifier is open source. Run the same check our web tool runs.

// open source · not on npm yet: clone the repository
import { verifyPhoto, summarizePhoto } from "@chainmail/verifier";

const bytes = await Bun.file("IMG_2048.JPG").bytes();
const report = verifyPhoto(bytes);
console.log(summarizePhoto(report));
// → Chainmail-verified: these bytes match the original stamp.
In development

Video and voice stamps

The same fingerprint for clips and voice notes.

In development · video and voice
How it works

Four steps.

1

Join the waitlist

Coming soon to the App Store. We email you when it is ready.

2

Turn on stamping

Link a Solana wallet, or use the one the app makes for you.

3

Send a photo

The stamp is sealed in before it leaves your phone.

4

They check it

Recipients see “Hashed on-chain”. Anyone can verify the file on the web.

Use cases

Where a stamp helps.

Friends and family groupsSelling something second-handLocal news tipsProof for a dispute
Security and privacy

Every result states its limits.

What a stamp proves
  • The JPEG has not changed since it was stamped.
  • The fingerprint matches a record on Solana, with its time. Devnet today; mainnet is in development.
  • Anyone can repeat the check with open-source code.
What it does not prove
  • That the photo was taken in Chainmail. Imported photos can be stamped too.
  • Who took the photo, or who sent it.
  • That a camera made it rather than an AI model.
  • That the scene was honest when the photo was taken.
Get started

Try the app.

Coming soon to the App Store. For teams, we set up a demo on your own photos.

See how we handle your data in our privacy policy.

Using photos as evidence at work? Book a demo.

See how we handle your data in our privacy policy.