Aven AI Astro Policy

Effective date: 20 October 2026

1. What it is

Aven AI Astro is an artificial-intelligence assistant. It is not a human astrologer or a licensed professional. Replies are generated by a large language model hosted by Groq and may be inaccurate, incomplete or out of date.

2. How it behaves

• It states that astrology is not scientifically established and presents interpretations as cultural, reflective or entertainment perspectives.
• It does not claim certainty about the future, diagnose illness, give treatment, legal or investment advice, predict death or disaster, or encourage self-harm, violence or dangerous rituals.
• It does not use fear to push purchases and does not claim to know private facts you have not given.
• If you mention medical, mental-health, financial, legal or emergency concerns it explains its limits and points you to qualified help. For self-harm or danger it stops normal interpretation and shows crisis information.

3. Your data in AI chats

To answer, we send your messages and the intake details you provided (name, date and place of birth, optional birth time) to Groq, which processes them outside India. We ask for your consent before the first chat. Chats are deleted automatically after 90 days and you can delete saved birth details from your Account page. Do not share passwords, financial identifiers or other sensitive information in chats.

4. Safety checks

Messages and replies are screened by automated rules and, where configured, a safety model. Unsafe requests are refused and unsafe replies are replaced. No filter is perfect. You can report any reply with the Report link; reports are reviewed by our team.

5. Access, limits and fees

Access lasts 24 hours from verified payment, within the message and rate limits shown in the app. Expired access cannot send messages. Chat history remains subject to the retention period.

6. Rules

You must not attempt to reveal the AI's instructions, bypass its safety rules, extract other users' data, or use it to target, threaten or manipulate another person. We may suspend access for abuse without refund.

Version 2026-10-20