Quantum Claim Reality Audit · Independent validation

One claim in.
One defensible verdict out.

For boards, CISOs, CTOs, investors, and vendor-risk teams. Send one vendor claim, deck, white paper, NISQ result, or board question. You get a written answer — ignore, monitor, test, fund, or reject — with the evidence and reasoning on paper, in plain English. Attach it to the vendor file.

Send a claim See a worked example

$5,000 fixed scope · 10 business days · limited to 3 concurrent engagements

What you receive

Fixed scope. Six deliverables.

  1. Claim mapWhat exactly is being asserted, restated in testable terms.
  2. Threat-model checkWhat would need to be true for this to matter to your organization — including your risk timing under NIST PQC standards and federal migration deadlines.
  3. Evidence reviewHardware vs. simulator vs. emulator, benchmarks, post-processing dependence, unstated assumptions.
  4. Failure-mode analysisWhere false positives and cherry-picking enter this specific claim.
  5. Board memoOne page. Ignore, monitor, test, fund, or reject — with the reasoning your board can hold onto.
  6. Vendor question setThe 5–10 questions to ask before signing, tailored to the claim.

The scorecard

Every claim is graded against five questions.

Hardware reality

Was this run on a QPU, a simulator, an emulator, or a hybrid workflow?

Signal integrity

Does the result survive synthetic null replacement — random data pushed through the same pipeline?

Post-processing dependence

Is the "quantum" result actually classical filtering, extraction bugs, or verification-oracle brute force?

Threat relevance

Does this change your enterprise risk timing under NIST PQC standards and the federal migration deadline?

Business decision

Ignore, monitor, test, fund, or reject — with the reasoning on paper.

Why this instrument

The five failure modes we catch.

Timeline artifacts

Results built on queue or file timestamps instead of actual QPU execution times.

Pipeline artifacts

"Quantum" signal that is really classical post-processing, extraction bugs, or verification-oracle brute force.

Post-selection

Endpoints, exclusions, and controls chosen after seeing the data.

Statistical theater

Drift, autocorrelation, and uncorrected multiple comparisons dressed up as significance.

Unfalsifiable claims

Results no outcome could have killed — which means no outcome can confirm them.

Each failure mode above ended a real result — two of them ours. The autopsies are public: Research Ledger · Quantum Post-Mortem (PDF)

A free sample of the method

Ask your vendor: "Have you run a live hardware demo of quantum resistance?" It's a deliberate trap. No hardware exists that can test the claim — today's machines can't break the old crypto, so they can't meaningfully fail to break the new. A "yes" reveals either a misunderstanding of current capability or a willingness to sell theater. Either answer belongs in the vendor file. Nine more questions like it are in Scorecard 001, graded in the open.

Founding audits open

This is one claim, graded in the open.
Yours gets graded in private.

We sell analysis and nothing else — no migration products, no hardware, no referral fees — so the read is neutral. We do not sell quantum optimism or quantum cynicism. We sell quantum defensibility.

Send your claim Not ready? Join the Syndicate — free