Faith and AI ethics do not usually share a headline, but for Lisa Kennell they are the same conversation. She spent years as a marketing executive for a financial firm, researching AI on the side before she said much about it out loud at work. Nobody on her team wanted it. She kept going anyway.
At a marketing conference in 2022, two months before ChatGPT existed, she scanned a QR code for a free AI tool and something clicked. She joined a mastermind, then another, and started building a baseline understanding while the tools around her kept changing shape. A year of prayer later, she left the corporate job and started Amplify Marketing, an agency built around one line she still won’t cross: AI helps her team work faster, it never replaces the people doing the work.
Lisa also writes a newsletter called Faith, Leadership, and AI, which puts her in a small group of people willing to talk about this out loud, in public, by name. On this episode of A Good Pour: Summer of Good AI, she and Kathryn talk about where that line actually sits, what happens when AI gets scripture wrong, and how to build an AI policy that protects what makes your work yours.
Why faith and AI ethics belong in the same room
Lisa’s reasoning starts with a simple fact about how AI tools work. Large language models learn from what people feed them. If people of faith stay quiet and stay off these tools, their voice does not stay neutral. It gets erased from the training data, or it gets represented by whoever did show up.
She has watched the reaction split by audience. Walk into a secular business and talk about AI, nobody blinks. Walk into a church or a ministry with the same conversation, and the room tenses. Lisa’s response to that tension is not to avoid it. It is to show up anyway, because someone has to help set the standards, and she would rather that be someone who takes scripture seriously than someone who does not.
Kathryn built a similar case from her own framework, CREATE, which starts from the belief that we are created to create. It is easy now to generate a picture of someone else and make them do anything. Consent still matters. Just because a tool makes something possible does not make it something you should do.
The AI sandwich: a simple check before anything goes out
Lisa has a name for the discipline that keeps AI in its place: the AI sandwich. An idea starts with a human. AI helps draft it, speed it up, or offer an angle you had not considered. Then a human checks it, every time, before it goes anywhere near an audience.
Kathryn’s own story shows why the third layer is not optional. While filling in for a youth pastor, she asked AI to help plan a Bible study on a passage in Isaiah. The tool explained what the scripture meant, and it was wrong. Not close, wrong. If she had trusted the answer and moved on, a room of teenagers would have walked away with a misreading of scripture that started as a shortcut.
That is the risk both Lisa and Kathryn keep circling back to: AI is confident whether or not it is right. It will hand you an answer with the same tone whether that answer is accurate or invented. The sandwich is not a bureaucratic step. It is the only thing standing between a shortcut and a mistake that goes out under your name.
Building an AI policy that turns faith and AI ethics into practice
When Lisa helps a business build its first AI policy, she does not start with a template. She starts with a question: what is this business known for, and how do you make sure that survives everything you automate?
From there, the policy gets specific. Which tasks can AI touch without a person reviewing first, and which ones always need a human check before they go out. How much autonomy an AI agent gets, whether it can send an email on its own or only draft one for someone to approve. Where the business draws its own version of Lisa’s line, because every business’s line will land in a slightly different place depending on what they do and who they serve.
The test Lisa comes back to again and again is not efficiency. It is whether the business still sounds like itself once AI is involved. A cloned voice that never gets a human check is efficient and hollow at the same time. The businesses she works with are not trying to do more with AI for its own sake. They are trying to protect the hours, and the humanity, that only a person can bring.
Use This Today
Before you publish or send anything AI helped you create this week, run it through Lisa’s version of the sandwich. Did the idea start with you? Did AI just help you move faster? Did you personally check it before anyone else saw it?
If any answer is no, that is the fix.
Resources
Connect with Lisa: amplify-marketing.com | LinkedIn: Lisa Kennell
Listen to the full episode here.