AI use

Beta. These pages have not been reviewed by counsel yet. No data protection officer is appointed for the free beta; write to privacy@lotusmind.one.

1. Where AI is used

• Gloop explanation card (after a session, verified members): an AI explains context. Every answer is labelled as an AI answer and carries a machine-readable marker (data-ai="1") on the text, so it is detectable as generated (EU AI Act Art. 50). It is not placed beside sacred text.
• Interface narration (optional, off by default): your browser's own speech engine reads interface text aloud. The voice is synthetic and is labelled as an AI voice. Narration passes interface strings only, never sacred text and never anything you typed. Depending on your browser and installed voices, that engine may use a remote service we do not control.
• Interface translations (Chinese, Spanish, Hindi, Vietnamese, Irish, Arabic, French): the interface strings in those languages were drafted by AI and have not been reviewed by a native speaker. This is stated here, in the Credits dialog, and marked in the locale files.

2. Where AI is never used

• No sacred text is written by AI. No scripture, prayer, ruling, fatwa, tafsir or blessing is generated, edited or translated by AI. Sacred text is shown verbatim from the sources under Credits and licences and checked by SHA-256 before it is displayed.
• The interface voice never reads sacred text. Recited audio is HELD until a licence exists.
• The bell in the Buddhism room is a synthesised sound, not a recording.

3. How Gloop is served, and what leaves the EEA

Gloop runs only on a model run by Lotus Mind (GLM-5.3-Flash), reached over our private network. No third-party AI provider receives your question: there is no fallback to another company, and this site's server refuses to start if one is configured. If our model is busy or unavailable, Gloop says "Gloop is busy, try again in a moment. Your question was not sent anywhere else." Your question (the app sends up to 300 characters), the tradition, the text reference and your language are sent only after you tick the consent box beside the question; without it, the server refuses the request. This site does not store or log the question or the answer; its log records only that an answer was given and how long it took. No children's data and no secrets are sent. The answer shows which model replied.

4. Limits

Gloop may be wrong. Gloop does not give religious advice, does not issue rulings, and does not try to convert anyone. It points you to the cited source instead. Answers are additionally checked on our server before display: an answer that reproduces a long verbatim passage of a staged sacred text, or that tries to issue a ruling or promise a religious outcome, is not shown.

5. Status

Beta. Not reviewed by counsel or by tradition advisers. Gloop is on for verified members, on our own model only. Turning on any third-party model would require a new data protection assessment first.

Controller: Lotus Mind Limited, company no. 826314, 10 Aderrig Manor, Adamstown, Lucan, Dublin K78 A4T8, Ireland · privacy@lotusmind.one · Complaints: Data Protection Commission, 21 Fitzwilliam Square South, Dublin 2, D02 RD28, Ireland — dataprotection.ie, 01 765 0100 / 1800 437 737.

Version lm-pray-dpia-20261111-v1 (beta). Translations of these pages were drafted by AI and checked only by AI reviewers, on the owner's decision; they are not legal advice and were not reviewed by counsel or a native speaker.