95Podcast 341 Summary – AI Guardrails for Churches: Practical Guidelines for Ethical, Safe Ministry Use (w/ Shelby Black)

95Podcast 341 Summary – AI Guardrails for Churches: Practical Guidelines for Ethical, Safe Ministry Use (w/ Shelby Black)

Dale Sellers and Joseph Bennett are joined by Shelby Black, co-founder of Bellwether, an AI-powered platform built specifically for ministry leaders, to discuss 10 practical AI guardrails for churches—data privacy, transparency, pastoral boundaries, automation risks, and how pastors can lead congregations already using AI.

Key Points In Brief

1 – AI assists. Humans Discern. AI is a tool that should assist humans—ministry leaders remain responsible for discernment and outcomes.

2 – Never input confidential information into public AI tools. Avoid putting confidential or personally identifying information into public/free AI tools; paid/enterprise settings and purpose-built tools are safer.

3 – AI cannot speak for God. AI cannot “speak for God” and should not be treated as spiritual authority or a replacement for theological discernment.

4 – Treat every AI output like a first draft. Verify, fact-check, and ask for citations to reduce hallucinations.

5 – Know what you’re agreeing to before you sign up. Read the fine print: know what you’re agreeing to (data retention, training, privacy settings).

6 – Be transparent with your congregation. Transparency matters: disclose meaningful AI use so trust isn’t damaged.

7 – Automate carefully. Mistakes multiply. Automate carefully with “human in the loop” oversight; don’t let automations run unchecked.

8 – AI cannot replace pastoral presence. AI cannot replace pastoral presence in crisis, grief, counseling, or spiritual care.

9 – Know who owns the tools and what happens when people leave. Assign ownership of AI tools, workflows, and prompts so processes don’t break when staff transitions happen.

10 – Train your congregation, not just your staff. Pastors should address AI openly because congregations are already using it for life, relationships, and even spiritual guidance.

Key Takeaways
  • Guardrails beat “rules.” Policies should protect people and trust without killing innovation.
  • Data stewardship is discipleship. Churches have a unique responsibility to protect member information and pastoral care notes.
  • Discernment stays human. AI can help refine communication, brainstorm, or summarize—but leaders must verify and own what’s shared.
  • Transparency prevents backlash. When AI use is hidden and later discovered, it can feel deceptive even if intentions were good.
  • Your congregation is already using AI. Ignoring it won’t stop it; shepherding people through it is part of cultural discipleship.
Notable Quotes
  • “If anyone claims they know everything about AI, they are lying—because it has probably changed since they started that sentence.” — Shelby Black
  • “AI cannot speak for God.” — Shelby Black
  • “Treat every AI output like a first draft.” — Shelby Black
  • “If you’re not comfortable claiming that it was AI, you’re probably not in a good place to use AI that much.” — Shelby Black
  • “They’re coming to Claude before they come to you for pastoral care.” — Shelby Black
  • “Show your work.” — Joseph Bennett (illustrating transparency in AI usage)
Next Steps (Practical for Pastors & Church Teams)
  1. Start low-stakes: Use a tool like Claude for a simple task (meal planning, drafting a basic email) to learn the workflow.
  2. Create a church AI “guardrails” doc: A 1-page guideline covering confidentiality, transparency, accuracy checks, and approvals.
  3. Draw a hard boundary on counseling notes: Do not paste pastoral care notes or counseling details into general AI tools.
  4. Decide your transparency threshold: For example, disclose AI involvement when it materially shaped content (graphics, curriculum, scripts, etc.).
  5. Assign an AI owner: Keep a shared list of AI tools, logins, prompts, and automations—and who is responsible for each.
  6. Teach your congregation: Consider a short sermon moment, class, or resource that addresses AI use and discernment.
Other 95Network AI Podcasts:

Link To Podcast Audio: 95Podcast 341

 

Link To Podcast YouTube:

Q & A Transcript

Q: Why does it feel “icky” to discover something was AI-generated?

A (Shelby Black): Because it can feel deceptive if people weren’t told upfront. Transparency builds trust, especially when AI is shaping content or communication.

Q: What’s a major safety rule with AI and church data?

A (Shelby Black): Don’t put confidential information into a public AI tool. Free tools may train on data; paid/enterprise options can offer better privacy controls. For sensitive workflows, use tools designed with data protection and masking.

Q: Can AI help pastors with sermon prep?

A (Joseph Bennett / discussion): It can assist (brainstorming, alternate perspectives, organization), but it must not become “the sermon.” Leaders need to check accuracy and theology—AI can fabricate or misquote sources.

Q: Why “AI cannot speak for God”?

A (Shelby Black): AI predicts what you want to hear based on patterns in human data. It’s not spiritual authority and can easily mirror human bias and preference rather than biblical truth.

Q: Why should pastors talk about AI publicly?

A (Shelby Black): People are already using AI for relationships, emotional health, and even spiritual questions. Pastors don’t need to know everything—just provide guidance and bring the conversation into the light.

 

Post & Podcast Categories

Posts & Podcasts by Month