Mira — AI & Community Content Policy

Covers Google Play's Generative AI requirements, defines the boundary for any future User-Generated Content (UGC) feature, and supports the brand's “reflection, not prediction” promise. The current release contains an AI companion and AI guide personas only—no live experts or user-to-user communication. Surfaced in-app and referenced from the store listing.

1. What Mira is — and is not

Mira offers reflection, guidance, and entertainment through astrology, tarot, and an AI companion. Mira does not predict the future and does not provide medical, mental-health, legal, financial, or safety advice. Every AI surface carries a visible "for reflection, not prediction" note.

2. Generative AI safeguards (Play Generative AI policy)

• Disclosure: the AI companion is clearly labeled as AI; users know they are not talking to a human.
• Prohibited outputs: the AI is instructed and filtered to refuse generating sexual content involving minors, content that facilitates self-harm, dangerous/illegal acts, harassment, hate, or deceptive health/financial/“guaranteed fate” claims.
• Input/Output moderation: user prompts and model outputs pass safety classification; unsafe content is blocked and logged.
• In-app reporting of AI content: every AI message can be reported in-app (Play requires a way to report offensive generative output). Reports are reviewed and feed prompt/guardrail tuning.
• No model training on personal chats where the provider exposes that control; minimize data sent.

3. Sensitive topics & crisis handling

• If a user expresses crisis, self-harm, abuse, or acute medical/mental-health distress, Mira does not give astrological "answers." It responds with care and surfaces region-appropriate professional and emergency resources (e.g., local crisis lines), and encourages contacting a qualified professional.
• Mira will not exploit fear or urgency, will not upsell during a disclosed crisis, and avoids deterministic claims about death, illness, pregnancy outcomes, or finances.

4. Current release has no live experts or user-to-user UGC

The area formerly labeled “Experts” contains AI guide personas only. They are AI interfaces, not people. The current release has no live expert marketplace, public feed, user profiles visible to others, or user-to-user messaging. Prompts are sent privately for AI processing and are not published to other users.

If Mira later adds human experts, community posting, or user-to-user communication, that feature must not be enabled until it has Play-compliant terms, in-context reporting, user blocking where applicable, moderation, account enforcement, and any required expert vetting. This future-feature boundary does not change the current release's no UGC sharing declaration.

5. Prohibited content & behavior (AI messages)

No content that is sexual involving minors; that harasses, threatens, or dehumanizes; that promotes self-harm, violence, or illegal acts; that constitutes fraud, scams, or deceptive claims; that impersonates others; or that violates privacy. Unsafe prompts or outputs may be blocked; repeated abuse can lead to account action.

6. Age

Mira is 18+. The Play Target audience & content declaration must exclude children, and onboarding/store presentation must not appeal to minors.

7. Enforcement & appeals

We review reports, take graduated action (warning → removal → suspension → ban), and offer a contact ([email protected]) to appeal decisions. We log moderation actions for accountability.

8. Compliance cross-references

• Play Generative AI app requirements · Play UGC policy · Play Health/Medical & Financial misrepresentation rules · FTC truth-in-advertising (objective claims must be substantiated).
• Keep this policy consistent with the Privacy Policy and the internal Data Safety mapping.