Responsible AI Use in Sermon Preparation

Keep responsibility with the preacher

Artificial intelligence can help a preacher organise notes, identify unclear sentences, generate discussion questions and explore alternative ways to explain an idea. It cannot assume responsibility for interpreting Scripture, discerning a congregation's needs or speaking truthfully in God's name. The person who preaches remains accountable for the message.

UNESCO's guidance on generative AI in education and research emphasises human agency, privacy and the evaluation of tools. Those principles are useful when considering ministry applications, although UNESCO is not a theological authority. The Christian task adds its own question: does this use help the preacher handle Scripture faithfully and love people responsibly?

Second Timothy 2:15 directs the minister towards approved work and the right handling of the word of truth. AI does not remove the discipline of study. A fluent paragraph should invite examination, not end it. The central issue is whether the preacher can understand, substantiate and responsibly communicate what will be said.

Study the passage before requesting a sermon

Begin with the biblical text itself. Read the surrounding passage, identify the literary form, consider the original audience and trace the argument. Write preliminary observations and questions before asking a tool to suggest an outline. This makes AI a support to your engagement rather than the source of your entire encounter with the passage.

Use a trusted Bible edition to verify every quotation. If preaching from the King James Version, check the words directly in that text. AI may blend translations, misidentify a reference or present a paraphrase as a quotation. A supplied Greek or Hebrew explanation also needs verification in a reliable language resource.

A useful request is: “Identify assumptions in my interpretation of this passage and suggest questions I should investigate.” The answer is a set of leads, not a verdict. Compare each proposal with the text and suitable scholarly resources. Record where interpreters disagree instead of treating the first confident answer as consensus.

Use AI for tasks you can evaluate

After forming your central interpretation, AI can help improve communication. Ask it to identify jargon, test whether an illustration supports the main point, suggest questions for group discussion or shorten an announcement. These tasks are easier to judge when you already know what the material should accomplish.

An illustrative example: a preacher preparing on the Good Samaritan writes an outline grounded in Luke 10. The preacher then asks AI to identify language that a teenager might not understand and to suggest three discussion questions about neighbourly responsibility. The preacher checks the suggestions and chooses what fits the actual congregation.

Do not allow useful formatting assistance to become unsupported authority. A tool may invent a quotation, research paper, historical detail or supposed personal testimony. Open and inspect the original source for any claim that matters. If the source cannot be found or does not support the claim, remove or replace it.

• Supply your own interpretation and intended audience before requesting editing help.
• Ask for uncertainty and alternative explanations, then verify them independently.
• Keep a short source record for important factual or historical claims.
• Read the finished message aloud and remove language you cannot explain naturally.
• Review application with a trusted colleague when the subject is sensitive.

Protect people and tell the truth

Pastoral preparation often includes confidential information. Do not paste identifiable counselling notes, private prayer requests, children's records or another person's sensitive story into a general AI service. Removing a name may be insufficient when the remaining details identify the person. Use a clearly fictional scenario where a teaching example is needed.

Never present an AI-generated event as something that happened to you. A fabricated conversion testimony or healing story does not become acceptable because it illustrates a valuable lesson. If an example is hypothetical, say so. If a story belongs to someone else, obtain appropriate permission and represent it accurately.

Follow your church, institution or publisher's expectations for acknowledging substantial AI assistance. Be especially clear where the audience could otherwise misunderstand the origin of a quotation, image, testimony or research claim. Honesty about process protects trust and allows others to judge the material appropriately.

Evaluate what the tool is doing to the preacher

Efficiency is not the only measure. Ask whether repeated use is strengthening interpretation, language and judgment, or making it easier to avoid them. A preacher who can produce more messages but is less able to explain the biblical argument has gained output while losing an important capacity.

Retain preparation practices that require direct engagement: reading the passage without generated commentary, praying, listening to people, consulting books, discussing difficult questions and revising your own words. A tool can assist these practices, but it should not displace the relationships and disciplines through which a minister grows.

Before preaching, apply four final questions: Is it faithful to the text? Are its factual claims supportable? Does it protect the people involved? Can I take responsibility for saying it? If any answer is uncertain, return to the material. Responsible AI use is the deliberate exercise of human judgment in the service of faithful ministry.

Sources and further reading

These sources inform the discussion. The practical examples and recommendations are explanatory applications, not claims that a named institution has implemented them.

1. UNESCO — Guidance for Generative AI in Education and Research
2. UNESCO — AI Must Be Kept in Check at School
3. Lausanne Movement — The Cape Town Commitment