AI & Automated Analysis Disclosure
Automation can organize evidence; it can also be wrong.
Review draft. This document has not been adopted as binding terms and is not legal advice.
Uses
Automation may classify equipment, normalize brands and models, find duplicates, identify comparables, estimate values, flag anomalies and summarize structured facts. Individual features depend on configured implementations; this page does not assert every capability is active.
Evidence and correction
Automated inference is not a verified fact. Review source records, confidence and material assumptions. Consequential corrections can be requested through /trust/report. Verification requires a separate evidence record and reviewer or trustworthy transaction evidence.
Community safeguards
AI-generated content must not impersonate real reviewers or invent customer experiences. Aggregated sentiment, individual opinion and KartExchange analysis must be labeled separately when those features are introduced.
Items for counsel and operator review
Review automated decision impacts, model/provider terms, data transfers, training uses and transparency requirements.
Version history
2026-09-05 — Initial engineering draft for counsel review; not adopted. Permanent version