AI Transparency Policy

Calia explains CRM-quality findings, answers product questions, prepares supported scan or deterministic correction proposals, and helps create credential-free support context. It is not a general autonomous CRM operator.

CleanIQ sends limited tenant context and recent conversation to the OpenAI API. It excludes stored connector credentials and full payment-card data, requests store: false, and instructs users not to submit secrets or restricted data in prompts.

Calia cannot bypass CleanIQ authorization, organization ownership, plan limits, provider constraints, or approval gates. Proposed actions are server-built, displayed for review, re-authorized at execution, and audited. High-risk deletions, merges, billing, access, credential, and administrative actions remain unavailable unless separately designed and approved.

AI output may be incomplete, outdated, biased, or incorrect. Users must review recommendations and remain responsible for decisions, provider rules, legal compliance, and the effect of approved actions. Calia does not provide professional advice.

Users can decline a proposal, use account-linked Support, or report an inaccurate or unsafe response. CleanIQ may review credential-free diagnostic context to investigate and improve and develop the service.