Verifiable AI

A machine's record, trusted like a person's.

AI trades now — sizes positions, picks entries, runs live capital — and it does it on nobody's proof. You are asked to trust a model's track record on the word of the people who built it, and you can't see inside the model to check. MIZAN closes that gap the same way it closes it for humans: the model's conduct becomes checkable without its weights ever being exposed. A compact model's inference is recomputed inside the proof; any model — up to frontier scale — proves its behaviour through the sealed path.

+895%
Return of an AI net
the gate REFUSED — on the wall
0
Weights, params, or prompts
that ever leave the machine
~81 ms
To verify a model's record,
offline, trusting no one
How long the credential
verifies — human or machine
§ 01The refusal is the credential

It refused an AI that returned +895%.

We put a neural network through the gate on NVDA. It returned +895% — and the gate refused it: a 38% drawdown and a Sharpe of 1.02, under the 1.20 bar. That verdict is published on the wall, credential 6b767764…, next to everything else.

That is the whole point. Anyone can claim their AI is honest. To our knowledge, nobody else can show you their own model, cryptographically examined, publicly failed, and left on the record. A gate that passes everything certifies nothing — as true for a machine as for its maker. When a model's record does pass, an allocator knows exactly what that is worth, because they can see what it refused.

AI made the fake cheap. One 2026 study generated 898 trading strategies for $40 — zero survived an honest backtest. The same machine that mass-produces plausible-looking track records is the reason a record now needs a proof no language model can fake. The proof that catches the fake is the proof that certifies the real.
§ 02Two paths, both sealed

Compact models prove inside. Frontier models prove behaviour.

In-circuit

Small enough to recompute

A compact model's inference is recomputed inside the zero-knowledge proof itself — the gate re-runs the model on the committed data and proves the result. The weights are an input the proof consumes and never reveals. What it decided is checkable; what it is stays sealed.

Sealed path

Too large to prove internally

A frontier-scale model can't be recomputed in-circuit — so it proves behaviourally: the model emits its committed positions, and the gate runs over that committed stream. The architecture, the parameters, the prompts never enter the room. Its conduct becomes a credential; its internals never leave the machine.

Either way, the record is minted to the model's operator, verifies forever against the era that judged it, and can be chained forward — a live pulse proving the model kept behaving, not just that it once backtested well. A self-improving agent's track record becomes a thing the world can check instead of take on faith.

§ 03The honest boundary

What a machine's credential does not claim.

The boundary is printed, as always. In-circuit inference today covers compact models; frontier-scale models are covered behaviourally via the sealed path, not internally. A sealed model cannot prove what it is not using — it proves what it did, not the absence of a given input. And no credential — human or machine — asserts future returns, capacity, or crowding. What it proves is conduct on committed data, honestly, against the true trial count. How the gate works → · the full boundary →

§ 04Prove a model

Bring the model you'd want an allocator to trust.

Mint a credential on what your model actually did — its weights never leave your machine, and the verdict is yours to publish or not. It may refuse you; it refused a net that returned +895%. Either way you learn, in a day, for free, exactly where your model stands. Minting is free while open testing lasts, and your number on the wall is permanent.

ENGINE v11 · 3ac3b10b… ● LIVE REGISTRY 77 CREDENTIALS · APPEND-ONLY VERIFY ~81 MS · OFFLINE · TRUSTING NO ONE ANCHOR BITCOIN #962,013 SPEC VTR-1 · FROZEN · CC BY ROOT 88298a2e…c6825a · MERKLE-COMMITTED