AI Transparency

The plain disclosure

The devotional reflections, prayers, and questions in Meno are written by artificial intelligence. They are generated for you, personally, by Anthropic’s Claude model, not written by a human pastor or author, and not reviewed by a human before you see them. Scripture itself is the exception: every Bible passage is the canonical text of a public-domain translation, supplied to the model verbatim so it cannot misquote it.

How a devotional is made

• The passage comes first, not from the AI. Your reading follows a structured two-year journey through the Bible (with seasonal readings on holidays). The AI does not choose what Scripture says or, on an ordinary day, which passage you read.
• Your context is minimized. We send the model a purpose-built summary of your profile and recent updates: only what helps it meet you where you are, not your raw records.
• Guardrails shape the writing. The model is instructed to stay grounded in the supplied passage, write within historic, orthodox Christian faith, avoid doctrinal overreach, and respond to crisis signals with care and real resources rather than a routine reading.
• Anthropic processes the minimized prompt.We use Anthropic's API data controls, but we have not yet obtained a contractual zero-retention arrangement. We will update this page when that additional protection is in place.
• We keep receipts. Every devotional records which model wrote it and what context shaped it, so “why did I get this?” always has an answer.

AI and spiritual guidance

This matters enough to say carefully: Meno is a companion to time in God’s Word; it is not a source of spiritual authority. An AI does not know God, does not pray, and does not carry the responsibility a pastor, elder, or wise friend carries for your soul. Its reflections can be genuinely helpful and still be wrong. Weigh everything it writes against Scripture itself and the counsel of your own church and community. If a devotional ever seems to conflict with Scripture, trust Scripture. And for the weightiest things (grief, crisis, doctrine, major decisions), bring people, not just an app.

Our responsible AI principles

• Scripture is supplied, never generated. The model may not quote from memory.
• Personalization serves you, never exploits you. Your story is used to meet you, not to profile, target, or sell to you.
• Minimum necessary data. The model sees a summary, not your records.
• Provider accountability. We disclose which AI provider processes a prompt and pursue the strongest retention and no-training terms available.
• Honesty about authorship. We will always tell you plainly that the writing is AI-generated.
• Care in crisis. Signals of acute distress are met with gentleness and real-world resources, not business as usual.

Related

What data the AI sees is covered in the Privacy Policy; your agreement about AI-generated content lives in the Terms of Use