The AI section of the security questionnaire
just got longer. Answer it once — hash-pinned.
73% of enterprise buyers gate on security attestations, and the AI sections of questionnaires now ask for model lineage, eval evidence, and drift monitoring. TrustPack assembles that evidence and pins it with a SHA-256 content hash, so reviewers can check it hasn’t changed since you sent it.
The free card is a structured, hash-pinned self-disclosure — every field labeled self-reported. Verified bundles are a paid engagement.
What AI questions now show up in security reviews and RFPs
Free: the AI Evidence Card
A clean, public, one-page card of your AI posture: models used, eval methodology, headline metrics, data-handling notes. Server-side we compute a SHA-256 hash of the canonical content and print it on the card — content-addressed, so any edit changes the hash. Link it from your trust center or paste it into a questionnaire response.
Honesty rule, stated up front: every field on a free card is visibly badged self-reported · unverified. The card is a structured, hash-pinned self-disclosure — AEQUARA has not verified the claims on it. What the hash gives a reviewer is integrity (it hasn’t changed), not verification (that it’s true). Verified evidence bundles are a paid engagement.
Founding TrustPack — verified, white-glove
For teams actively losing weeks to questionnaires: a human + AEQUARA’s substrate assemble a verified evidence bundle with you. This is a concierge engagement — onboarding starts within 2 business days of subscribing; it is not an instant automated export, and we don’t pretend otherwise.
Evidence inventory call
A working session to map what evidence you already have — evals, model inventories, data-flow docs — and what reviewers will ask for that you don’t.
Assembled evidence bundle
Your models, eval methodology, metrics, and data-handling posture assembled into a bundle mapped to the AI sections of common security questionnaires.
Hash-pinned + monthly refresh
The bundle is content-addressed with a SHA-256 hash, so a reviewer can confirm it hasn’t changed since you sent it — and refreshed monthly so it stays current.
Attestation letter
A letter from AEQUARA describing exactly what we reviewed, how, and when — scoped honestly to what was actually verified.
Founding tier — first 10 customers
price locked for founding customers · cancel anytime · monthly refresh included
Become a founding customer →TrustPack is built by AEQUARA — the calibrated-AI company. AEQUARA publishes a Model Calibration Index (ECE/Brier, SHA-256-pinned methodology) and a public Brier-tracked record: aequara.ai/trust. We hold our own claims to the standard we’re selling you.
TrustPack is not a certification, audit, or compliance framework, and an evidence bundle is not legal or compliance advice. Questions? hello@aequara.ai — a human reads it.