An AI interpretive framework is a named, reviewable set of instructions your church adopts for how AI should read sources, work through meaning, and prepare a response. Under the Storehouse proposal, it would make the church’s theological commitments, source authority, task method, uncertainty rules, output limits, and review requirements explicit.
Scripture would govern in its published, human-translated edition. The church’s adopted confession or statement of faith and recorded decisions would keep their identified standing. A sermon would show what its preacher taught; it would not, by itself, establish adopted doctrine. Permitted outside references would remain references, and AI’s explanation would carry no independent doctrinal authority.
The assistant would preserve context, distinguish source words from interpretation and proposed application, and ask for evidence or human judgment when sources conflict or leave gaps. It would not infer a church’s tradition from a sermon, generate Scripture, rewrite attributed speech, or expand access to private material.
For Wednesday's small group, these five discussion steps illustrate one task method within that broader framework:
- Connect. Start with a question about everyday life.
- Read. Ask what the passage and sermon actually say.
- Understand. Explore the sermon's explanation in context.
- Apply. Consider what that teaching means for this week.
- Respond. Invite a next step the group can discuss and take together.
The framework’s theological commitments, source rules, and limits would govern all five steps. The assistant would help prepare questions from the selected sermon, showing sources and flagging gaps. The leader checks the questions and guides the conversation.
For a supported task in the chat in testing, an admin adopts the guide before use and approves later changes. Storehouse can suggest revisions, but it cannot adopt them. A draft prepared with a method identifies that method; it still needs review before sharing.
Under the broader proposal, authorized church leadership would explicitly adopt an active AI interpretive framework for covered generation and approve changes before they take effect. Without adoption, the assistant would explain what is needed. Framework adoption would remain separate from approving each generated resource; finding sources, transcribing speech, and faithfully translating it would not depend on that adoption.
Explore AI frameworks and stewardship.
Chat in testing; unavailable in production. The broader AI interpretive framework and Transform uses described here remain proposed.