AI Transparency Notice
Last updated: May 2026
KnowMe uses AI to run the onboarding interview, evaluate compatibility, mediate conversations, and review content for safety. This notice explains how each AI system works and what rights you have.
AI systems in use
Interviewer (INTERVIEWER_V3) — structured onboarding, 12+ turns; Profile Builder — builds your latent compatibility profile; Matchmaker (MATCHMAKER_V2) — deterministic scoring + AI narrative, no agentic loop; Mediator — suggests conversation prompts; Safety Reviewer — screens messages and media; Termination Reflection (MEDIATOR_TERMINATION_V1) — optional one-time post-match reflection.
Fixed matching weights
Values alignment 35% · Relationship intent 30% · Communication style 25% · Safety score −10%. These weights are fixed in code and never change based on your subscription status or any commercial factor.
What the AI does not use
Your photo, demographic proxies for protected characteristics, your location (beyond country), social media profiles, or any third-party purchased data.
Special category data (Art. 9 GDPR)
The interviewer may explore topics related to religious beliefs, political opinions, mental health, and sexual orientation — only if you grant the optional Sensitive Categories Profiling consent. These signals are isolated and stripped before reaching the matchmaking AI.
Termination reflection
An optional one-time AI-generated private reflection when a match ends. Requires bilateral consent from both participants. Delivered silently in-app — no push, no email. Uses only your own responses, never your former match's.
Your rights
You can request an explanation of any match decision, a conceptual summary of your latent profile, or human review of a moderation decision. Contact privacy@gianlucamazza.it. Full details in the complete AI Transparency Notice.