No, Blackbaud, we’re not ready for Agentic AI

ByStephen Christopher Nill, JD and Steven L. Meyers, PhD

A few days ago, Blackbaud University held a webinar for the company’s nonprofit customers titled “Are You Ready for Agentic AI? A Human First Readiness Check for Nonprofits.” The session came one day after Blackbaud announced four new AI agents for its platform, so this was the education arm briefing practitioners the morning after a product launch.

What was said in that hour has disturbed us since. You can watch the replay or read the transcript we prepared.

The presenter, by his own description an educator rather than a salesperson, was plain about what this new class of software does. The new Blackbaud Development Agent “has agency,” he explained. “It’s doing its own thing with human oversight.”

He now had our full attention.

“It is going to be taking action,” he said. “That is its job. It is going to be reaching out to your donors.”

Had we heard that correctly?

He encouraged customers to give the agent a name, “because it does help with relatability.” The same presentation assured the audience that “the agents should never pretend to be a full person.”

From what we could tell during the webinar, the new Development Agent was well received. A school’s advancement director anticipated its impact “for further engagement, cultivation, stewardship, and solicitations,” and the presenter welcomed the comment: “I love that this agent is involved in every step of that donor life cycle.”

One of us (Meyers) asked from the floor whether the sudden wave of donor surveys he has been receiving means organizations are readying their records for AI. The answer allowed that it is “probably a mix of technology and training and timing.”

The hour had its careful moments. The data-readiness checklist included solicit codes and consent, the presenter pressed the point about people who should not be contacted at all, he repeated that agents do not replace people, and he offered, unprompted, the most important sentence of the session: “Just because we have the capacity to communicate in a certain way does not mean that we should.”

On that last point, we agree with him completely. It’s one thing for a software vendor to offer a product that tears down philanthropy and quite another for nonprofits to help it do so. We expect software companies to act in their own profit interests. But we expect nonprofit professionals to act for the preservation of philanthropy.

Every so-called safeguard described in the Blackbaud presentation governs how well the agent faces the donor without a human intermediary. Not one of them asks whether it should.

So we’re asking. Should a nonprofit organization ever turn loose an AI agent to communicate directly with donors and even solicit gifts from them?

Blackbaud just pivoted from AI at your side to empty-chair AI

After the webinar, we did some research. We learned that until just recently Blackbaud limited its AI to drafting a thank-you and handing the draft to a development officer, promising that “you retain the power to revise and personalize each individual message.” A person still wrote and sent every message. Holding Fire, the book one of us (Nill) has just published, calls that arrangement AI at your side, and we have no quarrel with it. In fact, the book’s whole purpose is teaching how to work ethically with AI at your side.

But we also learned that in March of this year, Blackbaud’s Raiser’s Edge NXT gave nonprofits the ability to remove the development officer from the picture by having an AI agent communicate directly with donors. The Development Agent, according to Blackbaud, “generates personalized email and SMS conversations, adding context and making asks at the right moment.”

Supervision is offered as an option: “the ability to review and approve communications before they’re sent.”

So now the human development officer is an option?

We call this empty-chair AI: the machine replaces the development officer by communicating with the donor directly.

On July 22 came the final domino, when Blackbaud released “fully autonomous donor engagement workflows under human supervision,” aimed at “every potential or dormant donor who is not in a major gift officer portfolio.” The webinar came the morning after.

When it comes to empty-chair AI, Blackbaud did not go first. The market has been dividing into the same two classes for years, and the vendors describe their own products plainly:

Disclose it or hide it, it harms relationships

An office that turns the agent on its donors inherits a decision Blackbaud leaves entirely to the organization: whether donors are told that the correspondence now comes from an AI.

Walk both roads with us, because we want you to understand that disclosure to donors does not solve the ethical challenge we are all faced with:

Suppose the organization discloses. In the Journal of Business Ethics this year, Zhao, Zhou, and Chen tested charitable appeals on 118,927 people and found that appeals labeled as AI-generated drew significantly fewer clicks. Donors read the label and concluded the organization was chasing efficiency rather than mission. They pulled back. The same comparison run on commercial companies found no penalty at all; the sincerity penalty falls on nonprofits alone.

The researchers found one setting where the penalty faded: when the charity explained that its AI works behind the scenes, cutting costs so that more help reaches the people it serves, willingness to give recovered. Read that framing again, because it describes AI at your side, a tool behind a person, not an AI machine doing the corresponding. Nothing in the study, and nothing published anywhere that we can find, tests whether any explanation rescues machine correspondence itself. What would the disclosure even say? Would it admit that the warm, personal letter was generated to save staff time? The honest road for empty-chair AI is measured to lose donors where it has been tested, and untested where its defenders hope it works.

Now suppose the organization decides not to warn donors that they are now communicating directly with an AI rather than with a human representative. In this scenario, donors on the list are meant to believe a person is writing to them. The warmth, the timing, the remembered gift, and the reply that knows what they said last month are the signals fundraisers have always used to tell donors that somebody here knows them, and the agent manufactures those signals on schedule, with the friendly name as the finishing touch.

The subterfuge holds only until discovery. The problem is that AI prose has well-known tells that donors are learning to recognize. Worse, the fact that growing numbers of nonprofits are having their AI agents communicate directly with donors is becoming general knowledge on its own. Once the donor realizes all those warm messages were manufactured, trust erodes, maybe completely.

Donors have already told the profession which road they fear more. When the sector’s own responsible-AI body asked 1,031 of them last year, the fear they ranked first was AI bots portrayed as humans representing a nonprofit, and donors who said an AI-enabled organization would get less from them outnumbered those who said more by 32 percent to 14.

Disclose it or hide it, the result is the same: eroded trust. Disclose it plainly, and donors pull back now. Dress it up in mission language, and you are running an experiment no researcher has run, on your own donors. Conceal it, and they pull back at discovery, harder still.

Worse, even organizations that do not employ empty-chair AI will get caught up in the loss of trust. That’s why we consider empty-chair AI history’s biggest ethical challenge to the fundraising profession.

The decision we nonprofit professionals must make

Blackbaud and certain other vendors have lost their way, but those of us who carry the torch for our nonprofit organizations must stay on the path that sustains donor trust. An office that turns the AI agent on and aims it at its own donors is harming philanthropy because doing so erodes the relationship of the donor to the organization, and in time the relationship of all donors to all organizations.

A working alternative standard exists. It fits in one sentence: a person reads and approves every communication a donor receives. Everything else AI offers a development office, including the research, the briefings, the drafting, and the Monday morning portfolio review, stays on the table.

Our profession’s codes of ethics do not yet expressly prohibit empty-chair AI. So we have written an open letter to AFP, CGP, AHP, CASE, the Giving Institute, and every other association, roundtable, and council that sets standards for our profession, asking them to write that one sentence into the codes. The letter is short, and it is open for you to add your signature.

If you are with us, read the letter and add your name at charitychannel.com/open-letter.

Authors

Stephen Christopher Nill, JD

Stephen Christopher Nill, JD, founded CharityChannel in 1992, building it into one of the largest online communities of nonprofit-sector practitioners in the world. He has edited, designed, and published dozens of books for the nonprofit sector through CharityChannel Press. Over four decades, Stephen has built, administered, and advised fundraising programs in the United States and internationally that have raised billions of dollars. He has served as senior vice president of development at a hospital system, chief advancement officer at a West Coast university, director of development at a parochial school he helped to found, and founder and executive director of a charity that feeds the homeless in Southern California.

The Grant Professionals Association presented him with its first President’s Award. The Grant Professionals Certification Institute later honored him with the Pauline G. Annarino Award for foundational contributions to its existence. Stephen has been a frequent commentator on charitable giving for CNN, Fox News, MSNBC (now MS NOW), and other news shows.

He has provided training to attorneys on advanced charitable gift planning techniques on behalf of the State Bar of California’s continuing legal education program. As a lawyer and consultant, he advised hundreds of nonprofit organizations and educational institutions.

Stephen’s engagement with artificial intelligence reaches back to the 1970s, decades before the current wave. Holding Fire: A Skeptic’s Framework for the Fundraiser with AI at Your Side draws on that long history and on his career inside development offices.

He writes weekly to development professionals at charitychannel.com. These days he’s busy writing his first novel and composing symphonic music in his music studio.

Steven L. Meyers, PhD

Steven L. Meyers, PhD, is the founder of the Center for Personalized Philanthropy at the American Committee for the Weizmann Institute of Science, where he served as vice president and a member of the management team for more than twenty years. Before founding the Center he was national director of planned giving at the same organization. His mantra, “the right gift, at the right time, for the right purpose, for the right donor,” has shaped a generation of gift planning practice.

His work has been cited by the Chronicle of Philanthropy and published in Planned Giving Today, the Journal of Gift Planning, Advancing Philanthropy, and Investments and Wealth Monitor. He is a contributing author for the Planned Giving Design Center and eJewishPhilanthropy.

He has spoken for the National Association of Charitable Gift Planners, the Association of Fundraising Professionals, Crescendo Interactive, State Street Global Advisors, and the Investment Management Consultants Association. He has served on philanthropy councils in New York and Connecticut, and on two national task forces developing guidelines for counting and reporting charitable gifts and measuring fundraising effectiveness.

In Personalized Philanthropy: Crash the Fundraising Matrix he gives the conventional structure a name, the Fundraising Matrix, in which gift officers are sorted into silos and donors are managed accordingly. His alternative designs each gift from the donor’s goals rather than the institution’s categories. CharityChannel Press published the second edition.

He holds a PhD from the University at Buffalo, a master’s in organization and management from Antioch University, and a BA from Knox College. His practice now focuses on coaching gift officers toward becoming enlightened generalists, helping them bring creativity and balance back to the work of building lasting philanthropic relationships.