Effective date: 20 April 2026 · Version: 2026.04-v1
What AI do we use?
Claude by Anthropic. Model: claude-sonnet-4-20250514.
What the AI does
Analyses your narrative to infer personality traits (Big Five OCEAN), values, communication style, and attachment style tendencies.
Scores compatibility with potential matches and generates human-readable reasoning.
Crafts personalised icebreakers rooted in both profiles.
Classifies messages for safety (harassment, hate speech, self-harm signals, spam).
Translates icebreakers between languages for cross-language matches.
What the AI does NOT do
Make clinical diagnoses. **All scores are AI inferences, not assessments.**
Autonomously ban users. Every moderation escalation is reviewed by a human.
Sell or share your narrative with advertisers.
Psychometric disclaimer
Your AI profile is a living inference based on your writing and interactions. It can be imperfect or biased. Always view it as a conversation starter about yourself, not a verdict.
Your rights
**Opt out of AI analysis** anytime in **Settings → Privacy → AI Profile**. Match quality will be reduced.
**Dispute an AI inference** if you feel it misrepresents you — we will remove or correct it.
**Request human review** of any AI decision affecting your account.