ComplyJet AI doesn't draft questionnaire answers out of nowhere — it draws from two libraries you build up under Prove → Trust → Questionnaires.
QA Library (Answer Library)
A reusable collection of question-and-answer pairs your org maintains once, then reuses across every incoming questionnaire.
Create Answer to add one manually, or approve items from a real questionnaire and add them here for reuse.
Each answer has an Owner (who's responsible for keeping it current — not necessarily who wrote it).
Resources
Source documents — SOC 2 reports, policies, security docs — that AI reads for context when drafting answers, rather than reusable Q&A pairs.
Create Resource to upload a document directly, or use Import from Policies to pull in your already-approved Policies in bulk (one-way — edits here don't flow back to the source policy).
The step people miss: "AI Readiness"
Adding something to either library does not make it usable by AI right away — notice the AI Readiness column. A new answer or resource starts out not ready; you have to explicitly Process for AI (individually, or Process All for AI in bulk) before it's actually indexed and available to questionnaire drafting. If AI-generated answers seem thin or missing context, this is usually why — check whether your library items actually show "Ready for AI" yet.