AI Transparency Statement

Flirty is made of software. This statement says what is generated, what automated decisions are taken, and what that means for you.

Everything here is AI. Every person you meet on Flirty, her face, her body, her photos and videos, and every message she writes are generated by artificial intelligence. No persona is a real person.

1. What is generated

1. The people. Personas are fictional characters. Their faces and bodies are generated images. Their names, histories and personalities are written by software and by our team.
2. Messages. Replies and the messages she starts herself are written by large language models, given her persona card, the recent conversation, short notes she keeps about you and her day.
3. Pictures and clips. Photos and short videos are generated by image and video models from text prompts and reference pictures of the persona.
4. Memory and notes. A model writes short notes about what you tell her. She also keeps private notes for herself about where the conversation is going.
5. There is no generated voice on Flirty today.

2. What that means for you

• She has no feelings, needs or life of her own. Warmth in a reply is generated text.
• She is not waiting for you and cannot be hurt by you leaving. If the app ever suggests otherwise, that is a defect: report it.
• She may say things that are wrong, contradictory or odd. Nothing she says is advice or a fact about the real world.
• Pictures can contain faults. They are images of a fictional person.
• Reality and fiction: do not treat what she says or shows as true of a real person.

3. Automated decisions

Software decides what she says, what she offers, what you see first and how a picture is priced. Software also refuses some requests and classifies pictures (see the Moderation Policy).

None of this is a decision that has legal or similarly significant effects on you in the sense of data-protection law. A person decides about closing an account and about reports and appeals. You can contest those decisions (see the Complaints and Content Removal Policy).

4. What we do not do

• We do not present generated content as a real person, a real event or a real photograph of anyone.
• We do not knowingly create content from the likeness of a real person.
• We do not sell your conversations or license them to others to train models.

5. Mistakes

Generated content can fail or be wrong. If a message or a picture fails to arrive after you paid, we return the tokens (see the Refund and Billing Policy).