false
OasisLMS
en,es
Login
Catalog
Who Decides? What 1,452 Blinded Fundraising Decisi ...
Recording
Recording
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Video Summary
The webinar examined who should control AI decisions in fundraising and philanthropy. A blind study compared recommendations from fundraising professionals and ChatGPT across 16 realistic scenarios, with more than 1,450 judgments from advancement colleagues. Overall, participants slightly favored human recommendations, but results varied by workflow: AI performed better in tasks such as prospect research, stewardship scheduling, and donor-trip planning, while people were preferred for ambiguous gift designations and decisions involving donor intent.<br /><br />The central finding was that neither humans nor AI should automatically be trusted to act. Participants favored human oversight in most cases, especially for consequential or donor-facing decisions. The presenters recommended assigning each workflow a clear level of AI authority—from informing or drafting to recommending or executing within defined rules—and naming a person responsible. They also encouraged bounded pilots with review criteria and clear stop conditions.<br /><br />Michelle Bates described Purdue for Life’s AI program, designed to amplify human connection while protecting data and using AI where it provides value. The organization is expanding tools already in its technology environment, providing staff training, and gathering feedback to guide future use. Its approach emphasizes responsible experimentation, staff involvement, and strengthening professional judgment rather than replacing it.
Keywords
artificial intelligence
fundraising
philanthropy
human oversight
ChatGPT
donor intent
prospect research
AI governance
×
Please select your language
1
English