AI Fundraising Build Pt. 8: Can AI Be Trusted with Donor Data?

Andrew Robison and Matt Bond discuss why AI tools can give different responses to the same prompt and how Catholic fundraisers can work with that variation instead of fighting it. They explain how large language models generate answers by interpreting patterns rather than copying and pasting, why “different” usually isn’t “wrong,” and how to distinguish normal variation from true hallucinations. They also share practical ways to improve consistency so that outputs support sound fundraising strategy. 00:00 Intro02:50 Why AI Answers Vary04:04 LLMs Explained Simply08:18 Not Hallucinations, but Creativity11:52 When Strategy Must Be Precise13:48 Leveling Up Prompting Skills17:37 Practical Takeaways and Mindset

Andrew Robison and Matt Bond continue Petrus Development's “Using AI to Build a Development Program” demo series by working with donor data from the fictional St. Jude Catholic Leadership Center to show how an AI tool can help create a segmentation strategy. Andrew and Matt discuss prioritizing practical segments, “trust but verify” when AI hallucinates details, and the fundraiser’s role in selecting what’s useful. 00:00 Intro04:26 St Jude Data Overview05:32 What Segmentation Is09:54 Demo Spreadsheet and Prompt12:03 Pete Suggests RFG Segments17:15 December Mailing Segments20:56 Picking Top Segments22:08 Hallucinations: Trust/Verify24:37 How Pete Is Trained26:17 Key Takeaways and Next Steps

Sarah and Jenni from Petrus Development discuss whether relying on AI can make Catholic fundraisers “lazy,” especially in the skill-building work of writing, planning, and critical thinking. They explain that while tools like Pete AI can speed up tedious tasks and reduce errors, they can also tempt users to outsource the mental work that develops strategy, creativity, and confidence, or replace collaboration with coworkers. 00:00 Intro03:34 Will AI Make Us Lazy03:45 Writing and Planning Skills06:01 Brain Plasticity and Collaboration09:44 Finding the Line12:02 Prayerful Reflection13:30 Complacency - Copy Paste Trap15:09 Confidence and AI Insecurity18:22 Keeping AI in Order19:23 Wrap Up and Next Episode

Jenni Reamer and Sarah Rose review and improve the St. Jude Catholic Leadership Center’s fictional online giving page using AI. They explain that typical nonprofit giving-page conversion rates average around 10%, while optimized forms can reach 20–30%, and note that many organizations’ pages are not optimized. They then begin a deeper look at how to set suggested gift amounts based on typical gift size and real program costs, and preview upcoming episodes on segmentation and concerns about AI making fundraisers lazy.00:00 Intro03:22 Why Giving Pages Matter05:19 Conversion Rates and Optimization07:13 Tour of St Jude Giving Page09:10 Initial Critique Text and Friction10:55 AI Audit Top Fixes16:07 Deep Dive Suggested Gift Amounts20:13 Wrap Up

Matt Bond and Andrew Robison of Petrus Development reflect on how AI is affecting Catholic fundraising work so far as they build a fictitious development program from scratch. They discuss where AI has already helped, including drafting a case for support and speeding up administrative and data tasks, while noting that its impact on donor-facing outreach remains to be seen. They propose evaluating AI’s true value by results like more donor meetings, improved ROI, larger average gifts, and stronger retention, and preview an upcoming episode analyzing a website giving page.00:00 Show Intro and Setup00:53 Series Goal and Format03:25 Is AI Making Us Better05:42 Red Teaming with AI09:08 When AI Hurts Productivity11:26 Measuring Fundraising Impact13:33 Wrap Up and Goodbye

Andrew Robison and Matt Bond welcome viewers to a Tuesday demo episode of the Catholic Tech Table, showing how AI can help build a first-year fundraising budget for the fictional St. Jude Catholic Leadership Center as part of a 20-week series on creating a sustainable development program. 00:00 Series Recap and Budget Focus05:53 Budgeting Stakeholders Tip06:38 Demo Setup St Jude Fact Sheet09:00 Finding a First Year Goal12:15 Stress Testing Assumptions16:02 What If We Hire Staff22:24 Budget Expenses Breakdown29:26 Wrap Up and Next Episodes

In episode 10 of Catholic Tech Table’s series on using AI for Catholic fundraising, Andrew Robison and Matt Bond discuss AI “hallucinations”, the confident-sounding answers that are incorrect or fabricated, and why they happen. Using a real example of a made-up Harvard Business Review statistic, they explain that AI often generates likely responses based on patterns rather than verified truth, making errors especially risky when asking for specific facts, statistics, grants, deadlines, or quotes. 00:00 AI Hallucinations Intro03:47 Recap And Core Question04:25 What Is A Hallucination06:08 HBR Hallucination Story10:29 Why Hallucinations Happen14:21 Fundraising Risk Areas15:40 How To Verify Outputs19:24 Embarrassing Date Errors20:58 Wrap Up

Andrew Robison and Matt Bond demo how to use AI to design a repeatable workflow for cleaning messy spreadsheet data before importing into a CRM, using the fictional St. Jude Catholic Leadership Center as an example. They emphasize using AI to think through processes and speed work without exposing sensitive donor information, advocating strong donor privacy practices. 00:00 Intro01:08 Series Recap And Pete AI02:33 Sponsor: The Harbor03:39 St Jude Data Problem05:38 Data Horror Stories08:03 Privacy And Safe AI Use08:51 What To Clean Before Import11:10 Demo Prompt - Top Five Tasks16:08 Systematize Rules And Salutations19:18 Next Episodes And Wrap Up

Just because AI can process donor data does not mean it should.In this episode of the Catholic Tech Table’s Fundraising Build series, we continue examining how St. Jude Catholic Leadership Center can organize a decade of messy donor records while protecting the privacy of the people who entrusted that information to the organization.We discuss what should never be entered into a general AI tool. We also explore how fundraisers can still use AI safely.The central lesson is simple: AI can be a powerful assistant, but fundraisers remain responsible for guarding donor information.

How do you begin fundraising when donor and participant records are scattered across spreadsheets, registration forms, email lists, and old files?In this episode of the Catholic Tech Table’s Fundraising Build series, we return to St. Jude Catholic Leadership Center, where 10 years of inconsistent record keeping have left the organization without a centralized donor database or CRM.We also address an essential question: how can fundraisers use AI to organize data without exposing sensitive donor information?This episode establishes the structure and safeguards St. Jude needs before cleaning and importing its records into a CRM.