Legal
AI Disclosure
Where we use automation, where we do not, and what a person always decides.
Last updated 19 November 2026
People decide who joins
Every application is reviewed individually, by a person, not an algorithm. No automated system accepts, waitlists or declines anyone. Identity documents are also checked by a person.
How matches are suggested
Introductions start from a compatibility score, worked out from your answers, preferences and deal-breakers by rules we wrote. It is not generative AI. The strongest suggestions are introduced automatically; the rest are reviewed by a person on our team, who approves or holds each one.
Other automated checks
- Report screening. Report text is checked against a list of serious terms. A match makes the report urgent and alerts our team; it never closes or dismisses a report. A person reviews every report.
- Screenshots. The app detects screenshots. The first brings a warning, and the third bans the account automatically.
- Duplicate accounts. Device checks stop one person holding several applications.
- Uploads. Every photo and document is checked by its contents to confirm it is a genuine image.
If an automated step affected your account, write to privacy@plausone.com and a person will review it.
AI features that are built but switched off
The app includes a few AI features that are not switched on today. Before we switch any of them on, we will update this page and the Privacy Policy.
- Tone check (Anthropic’s Claude): a private suggestion, shown only to you, if a bio, report or support message you are about to send reads as hostile. The text is not stored, and moderators never see it.
- Concierge suggestions (Claude), for Reserved and The Circle members: ideas for events and when to get back in touch, based on first names, cities, event names and timing.
- Voice-note transcription (OpenAI): only for a voice note you choose to transcribe. The audio is sent for that one transcription and then discarded.
- Translation (DeepL): sent from your own phone, only for a message you choose to translate.
What AI never does here
- AI never reads your private messages. They are end-to-end encrypted, so our servers cannot read them either.
- AI never decides your application, and never removes a member on its own.
- We do not use your information to train AI models.
Questions about automation or AI, or to ask a person to review an automated decision: privacy@plausone.com.