Inside Philanthropy, like many publications, has been tracking the discussion around AI since the technology went mainstream with ChatGPT and other platforms. Paralleling big voices in technology and public policy at large, philanthropies are weighing in about AI’s possible benefits as well as potential dangers. But alongside that discussion, there’s also the question of how the people working in philanthropic organizations can best integrate AI technology in the service of their overall mission and everyday work.
Anyone who has worked in a large business or organization has likely experienced the arrival of new technologies. Maybe it’s a new computer system or software solution that’s going to help everyone do their job better and boost productivity. Sometimes they help and sometimes they hurt. Now, with the steep and sudden spread of AI, leaders in grantmaking organizations face the same kinds of questions that management in business and other sectors must address: how to ensure that this powerful and protean technology can help them achieve their goals, safely.
Those are the sorts of questions on the mind of Cinny Kennard, executive director of the Los Angeles-based Annenberg Foundation. One of the nation’s larger family foundations, with total assets of $1.47 billion as of 2019, Annenberg supports nonprofits in communities in California, throughout the U.S. and globally. Its grantmaking and other work focuses on a wide range of causes, including education, the arts, public policy, community needs, nature and wildlife, journalism and technology, among others. That means the Annenberg grantmaking team conducts a lot of research in a lot of communities and countries — and AI is increasingly in the middle of it all.
“I had all these different departments that were being given information and education about AI, and how it was going to apply in those departments,” Kennard said. But there was no real coordination across the organization, and that raised a lot of concerns. “What are the vulnerabilities around AI?” she said. “What are our data governance policies? What are the guardrails here? That has to be a conversation in philanthropic organizations.”
Kennard knows she’s not alone in these concerns, having talked to numerous colleagues and leadership counterparts in philanthropy who have similar questions. “I am not an expert in AI, but I sure have a responsibility to this organization to inform and educate all of us internally,” she said. “We’re managing billions of dollars that are supposed to be addressed to the greater good. Can AI help us do that? I don’t know the answer to that question.”
Expert advice
To begin to answer those questions about AI and how it could be used at Annenberg and throughout the philanthropic sector, Kennard turned to an expert. Last month, the foundation brought in technology strategist and social impact entrepreneur Chantal Forster, who will work with the Annenberg team during a six-month residency. Forster spent six years as executive director of the Technology Association of Grantmakers (TAG), which was established to find ways to use information technology to advance philanthropy’s goals. Not surprisingly, AI is top of mind for TAG.
Forster was most recently Special Adviser on AI at TAG, and had already begun a deep dive into the potential role of AI within philanthropic foundations, where she encountered a hard-to-manage mix of optimism from those who thought AI would solve every problem, and fear that it would trigger the apocalypse. She also saw that AI was already falling into a pattern that affects nearly every organization using tech at scale: IT’s potential disconnection from business decision-making.
“Historically, at most foundations, there’s a gap between the program teams who control the philanthropic purse and the operations teams who possess more expertise in the strategic use of technology,” Forster said. Further complicating the AI picture in philanthropy, as in many other sectors, is the direct-to-consumer way that AI apps are marketed. “They bypass the IT manager or director, and the IT director no longer has control over people’s use of the AI technology,” Forster said. This can lead to exactly the kind of lack of consistent policies that Annenberg’s Kennard fears.
At the Technology Association of Grantmakers, Forster coached philanthropic IT directors and chief information officers to get more involved in program goals. “I’ve spent my whole career in technology, but I’ve always been on the human side,” Forster said. “When I think about the use of technology, I think about it as a means to an end, not the end in itself. I have met with every single staff member at the Annenberg Foundation, and I didn’t lead with, ‘How do you want to use AI?’ I led with, ‘Tell me about your work. What’s challenging for you? What would you like to be easier? What would you like to do that you can’t currently do?’”
Top-down planning
At Annenberg, as at many foundations, AI has numerous potential uses, from the individual to the organizational level. Individuals might use AI to draft emails, memos, reports or other documents, or make transcriptions. At the organizational level, AI tools might manage, summarize or synthesize grants or grantee data. AI could also be used to support program development or grantmaking decisions, or to landscape issue areas or demographics.
For Forster at Annenberg, recommending or developing technology solutions will come later, only after she has a better grasp of how the people who work at the foundation and manage programs conduct their work. What is not clear is what risks an evolving technology like AI — already known for its penchant for alternative facts and concerns around data privacy — will adhere to Annenberg’s, or any grantmaker’s, organizational standards and values.
As Annenberg considers AI’s potential uses, it’ll be expanding on guidelines in a report, “Responsible AI Adoption in Philanthropy: An Initial Framework for Grantmakers,” published late last year in partnership with Project Evident and TAG. The publication, informed by input from 300 professionals working in the charitable and social sector, contains guidelines for the adoption of AI, including checklists and questions philanthropy leaders and managers need to ask as they start to use AI-based tools.
Now, a month and a half into her residency at Annenberg, Forster stresses the need for buy-in by top management. “Most foundations don’t move this quickly about technology,” she said. “This is not an initiative that’s just housed in the IT department, this is holistic wrestling with AI at the foundation.” That could be the most important takeaway from Annenberg’s assessment of AI: that senior management must ensure that technology is integrated deliberately, with a broader strategy and policy — and not in a haphazard or piecemeal fashion that could create problems or conflict with funders’ broader strategy and mission.
