trustless-ai trustless-ai
About & Press

Infrastructure you never have to trust — only recompute.

trustless-ai is an open ecosystem of standards, reference implementations and tooling for AI agents: infrastructure anyone can provide, anyone can use, and everyone can verify by recomputing the proof. It is owned by no one.

What it stands for

Don't trust — recompute.

Nothing ships that a third party can't independently verify from public data. Every claim is re-derived from the primary artifact, pinned to an exact reference, and answered with a tri-state verdict — verified-good, verified-bad, or unverifiable. "Couldn't check" is its own answer, never a silent pass. There is no pitch here; the substance is the pitch.

Verify it yourself

Don't take this page's word for any of it — re-run it:

The standards

The agent-standards family — each an ERC or a recomputable reference implementation:

Founders

Contributors

Each person authors the layer they own; coordination is by hand across channels, and the organization is the shared, CC0 home.

Contact & press

The work lives in the open on github.com/trustless-ai. For anything specific, the fastest path is an issue on the relevant repository — where it can be checked against the artifact, like everything else.