Dear Ethicist,

A vice president at my organization recently attended a conference presentation on using artificial intelligence (AI) to optimize donor segmentation, and we are now implementing a new AI driven project. While there is a general sense of enthusiasm for using new technology to increase our fundraising results, a few folks have been asking questions about the ethical implications of AI and are wary about the project.

One colleague’s comment about the AI project was, “I’m not sure how AI would work in fundraising since it seems like something that is so personal and relationship-based. Maybe it can be used for data segmentation but I have concerns about how it would work with donor interactions.”

Our vice president feels that the colleague is being reluctant to embrace change because if a vendor is selling a tool that uses AI, it must be ok, since it is now a part of fundraising. We are just using data, what’s the harm in that?

Sincerely,

Ready for the Future

Dear Ready,

We encounter AI in our everyday lives — it runs our phones, navigates our vehicles, determines the ads we see, etc. AI feels like the exciting buzzword of the day. While many seem eager to adopt it, few people (in my humble opinion) invest time in developing a deeper understanding of what it is, how it works and/or how it can be effectively and appropriately used. It also feels like claims about AI’s efficacy and benefits are frequently taken at face value, with little being done post-implementation to measure and test these claims.

One of the biggest opportunities that AI offers fundraising is the ability to analyze data and patterns in past behavior to make predictions about future behavior (giving). However, we must stop and consider how the data used here is collected. Is the data ethically sourced? Are there inherent biases in the data that could impact the results?

When AI is fueled by biased data, it can lead to bad predictions, discrimination and reinforcement of racist and sexist systems. The Vendor Due Diligence Toolkit can help you as you select a vendor for an AI project.

Before implementing an AI based project, here are some additional questions to consider:

• What is the business purpose of your organization using AI?
• What are the benefits and limitations of the AI project?
• What data of your own will the project be utilizing? Will your data be shared with the vendor?
• How will you effectively operationalize the project?
• How will you measure and report on the project’s return on investment?
• Have you considered any local/national/international privacy laws that could impact your project?

There are legitimate concerns that your colleagues may have with AI. One concern includes creating new processes with a “set it and forget it” mentality. Constant oversight and built-in evolution need to be in place to ensure AI is delivering on the promise of improved donor experience and targeting, and not creating a new set of problems that provide a poor donor experience. AI is not inherently creative or ethical, making human intelligence irreplaceable. AI can’t replace the social and contextual intelligence that humans provide.

One way to navigate ethical concerns here is to educate your entire team on AI. Having a base understanding of AI will allow your organization to understand the ethical concerns and make the right decisions. Allison Fine and Beth Kanter released a 2021 report on the current opportunities and challenges presented by AI in philanthropy. AI for Good Foundation has resources on setting standards for AI ethics best practices. An April 2022 blog from Carnegie Mellon’s Arts Management & Technology Lab offers an examination with “Navigating the Ethics of Using AI for Donor Solicitation.”

As Eleanor Roosevelt said, "The future belongs to those who believe in the beauty of their dreams.” Dream big, but make sure you stay rooted in ethical best practices.

Cheers!

The Ethicist