A governed answer library
Every answer is paired with an owner, source artifact, approval state, and an explicit unknown when evidence is missing.
AI-CAIQ V1.1 EVIDENCE FACTORY FOR ENTERPRISE AI VENDORS
AssureDeck turns scattered policies, model notes, eval results, and engineering facts into buyer-ready answers with evidence behind every claim—without pretending a template is proof.
CSA released AI-CAIQ v1.1 on June 23, 2026. Its companion AICM v1.1 contains 247 controls across 18 domains.
Read the CSA sourceAI PROCUREMENT PACKET
Model & provider inventory
OWNEREDCustomer data-use statement
APPROVEDEvaluation thresholds & results
2 GAPSAI incident response exercise
UNPROVENFREE / LOCAL / FOUR MINUTES
Runs in this browser. Nothing is uploaded, stored, or sent to AssureDeck.
This original 20-signal check is an informational pre-assessment. It does not reproduce or replace the official AI-CAIQ, and it is not an audit, certification, assurance opinion, or legal assessment.
WHAT THE BUYER CAN ACTUALLY REVIEW
The deliverable is designed to survive security, procurement, product, and legal review because every answer exposes its source and approval boundary.
Every answer is paired with an owner, source artifact, approval state, and an explicit unknown when evidence is missing.
Models, providers, data flows, intended uses, limitations, and human checkpoints become one reviewable system record.
Missing proof becomes a ranked engineering, policy, or operating task—not a vague red cell in a spreadsheet.
Approved material is packaged for your trust center, GRC workspace, buyer questionnaire, and STAR for AI pathway.
THE EVIDENCE FACTORY SPRINT
Built for AI-native SaaS teams already selling into enterprise accounts. One system boundary, one accountable client team, one reusable response library.
Inventory systems, providers, enterprise promises, buyer questions, and the evidence you already have.
Map evidence to the AI-CAIQ response set, mark unsupported claims, and assign owners to every material gap.
Draft missing artifacts, improve technical controls where scoped, and produce source-linked response language.
Run executive review, freeze approved answers, hand off imports, and rehearse the next buyer review.
FIXED-SCOPE ENGAGEMENTS
Fixed prices cover one defined AI system. Work starts after the agreed kickoff inputs are complete. Scope changes are priced before work, and no payment is taken on this site.
ENTRY
$5,000 / fixed
BUILD
$25,000 / fixed
OPERATE
$3,000 / month
DIRECT / CONFIDENTIALITY-AWARE
Send only the business context needed to assess fit. We review every request manually and confirm scope, exclusions, timeline, and fixed price before any work or payment begins.
OPEN METHODS / PRIMARY SOURCES
Practical guides explain how scope, claims, evidence, owners, and gaps fit together. Every framework fact links to its official source. The working inventory is free with no email gate.
Illustrative method only. The claim is not about AssureDeck or a customer and is not evidence until a client validates its own sources.
Run the no-signup check locally in your browser and export a Markdown gap report.
Run the free check ->01 / AI-CAIQ V1.1Scope one system, map answers to owners and sources, and keep partial or missing support visible.
Read the guide ->02 / BUYER REVIEWConvert a live buyer spreadsheet into a reusable answer library and evidence index.
Open the checklist ->03 / STAR FOR AIUnderstand CSA's current level path and separate preparation support from program decisions.
Check readiness ->04 / FIXED SPRINTReview the system card, claim register, evidence index, gap backlog, scope, and exclusions.
See the deliverables ->FREE / NO EMAIL GATE
Original working columns for scope, owner, approved answer, evidence URI, review date, gap, and target date.
BOUNDARIES MATTER
No. AssureDeck is an evidence-production and deal-enablement service. Your team approves every representation. Independent certification, assurance, or legal opinions stay with qualified third parties.
No. We make those systems more useful by creating the AI-specific inventory, evidence index, approved answers, and gap backlog that can be imported into the tools you already use.
That is expected. We never invent a control or soften an unknown. We separate what is proven, what is partially supported, and what must be built before the claim can be approved.
No. The readiness check runs locally in your browser. It asks only whether broad evidence types exist and does not upload or retain your answers.
The contact form stores only the fields you submit plus a one-way daily anti-abuse identifier. It is used to assess and respond to your request, is not sold, and is separate from the local-only readiness check. Use the Privacy request option in the same form for access or deletion.
THE BUYER IS ASKING NOW