Navigating AI complexity: Framing the conversation for philanthropy

Michael Pratt and Jason Franklin of Ktisis Capital provide a framework for discussing and evaluating AI in the context of philanthropy.

The generative artificial intelligence (AI) boom is sweeping across industries and sectors. Philanthropy is no exception, evidenced by much writing, research, and thought leadership on "AI and philanthropy." But AI and philanthropy can and does mean many things—including enhancing organizational efficiencies, innovating new approaches to social problems, reimagining long-standing sector functions, grappling with ethical concerns and considerations, exploring policy options, and more. The corresponding range of resources reflects that multitude of meanings and applications—and navigating it can feel mountainous.

Part of the issue is the words themselves. While most recent conversations have surrounded generative or creative AI, various technologies fall under the umbrella term. A common distinction is made between narrow AI and general AI. General AI is shorthand for artificial general intelligence, also called strong AI, and typically refers to technologies that match or surpass general human cognitive ability. Many consider generative or creative AI, like OpenAI's GPT-4 or Google's Gemini, to be the next generation of technologies within the narrow AI subset. Generative or creative AI is characterized by its ability to generate or create new data or information. Conversely, narrow AI, also called traditional or weak AI, typically refers to technologies that are better than humans at specific, narrowly defined tasks. Longer-standing narrow AI technologies have primarily focused on making predictions about a specific dataset that would take months or years with traditional human-powered analysis.

Further adding to the confusion is that the term "philanthropy" has long resisted orderly definition. Are we talking about the philanthropic sector as a whole? Nonprofit and community organizations pursuing impact? Individual donors and fundraising ecosystems? Large private foundations and other grantmakers? All of the above? Philanthropy is not a monolith but is often perceived and treated as one—further compounding the conversation's messiness.

Lastly, conversations about AI and philanthropy can get confusing because they often conflate or ignore the intended level of analysis or inquiry. When we talk about AI and philanthropy, are we focusing on grantees, grantmakers, individual donors, or the whole sector? Clarifying our level of analysis is essential, as it defines the scope of our conversation, which can vary wildly given the incredible diversity of implicated individuals, groups, and entities. Critically, we must also consider how the voices, perspectives, and lived experiences of those most proximate to problems show up in these conversations. While imperfect and not exhaustive, the list below may be a helpful starting point for clarifying the level of analysis of the various AI and philanthropy conversations you are currently having or inevitably find yourself in.

How does AI impact:

• Nonprofits as organizations, teams, individual employees, and volunteers?
• Grantmakers and individual donors, their teams, and their decision-making processes?
• Communities as complex ecosystems, the coalitions working in them, and the neighborhoods and residents who make up a community?
• Society in terms of the relationships between the philanthropic sector and government, the policy environment for AI and how it impacts society, and the impacts of AI on the marketplace and economy that will define and change the philanthropic sector's work?

Once we focus our level of analysis, we're better positioned to ask critical questions—and draw a line from those questions to important implications, urgent concerns, best next actions, and more. At Ktisis Capital, we've been using this framing to navigate our own understanding and thinking about AI and AI issues. Through those conversations—and trying our best to keep up with the pileup of resources—some reflections and observations that feel critical to consider include:

• Philanthropy and AI policy. There are many compelling calls for philanthropy to play a role in responsible AI adoption in the face of growing ethical and equity concerns. However, there seems to be far less talk and writing about the specific role philanthropy should play in shaping AI policy—and the specifics of the current debates, trends, and shifts in the AI policy space. Philanthropy must be prepared to engage in the policy conversation at both federal and state levels—foundations and donors must fund critical research, public education, lobbying, and other work to ensure that ethical and equitable AI policies emerge in the future.
• AI funding trends. Foundations and donors must educate themselves about the role of AI in society and act on opportunities to bring these considerations into their current funding. Additionally, industry-connected funders must tread cautiously and consider the ethical dynamics of their own engagement in the field.
• Digital divides and socio-economic disparities. The digital age is often framed in terms of the ways technological advancements have led to significant social changes and progress. However, a critical review of its history makes it clear that progress has had an unequal impact, with the economic benefits of innovations mainly accruing to a very small number of wealthy elites. The increasingly wide digital divide that existed among the general population preceding the AI era does not bode well for future equitable outcomes unless we proactively intervene as a society.
• Philanthropy's history with emerging technologies. Given philanthropy's comparable insulation from market conditions and forces, it has historically been slow to adopt or respond to tectonic social, economic, and technological shifts. That is proving out with AI technologies in concerning ways—at both the institutional and sector levels. There are terrific resources addressing the shortfalls and urgency of this, but more effort and resources are needed. Philanthropic leaders, from nonprofits to funders, must educate themselves quickly if they are to show up well in the critical AI conversations that will impact our society for the next century.
• Equity-centered AI adoption at the organizational level. On their current path, AI tools will replicate human biases similarly to past digital innovations and analyses—what will we do to shift course? How can we ensure robust, diverse data inputs to prevent or minimize discriminatory results? Which communities and leaders will emerge as AI pioneers, and how will we ensure access to AI capabilities for people from lower-income and other marginalized communities who often benefit least from technological developments?

Grappling with the impact of AI on philanthropy and broader society will be a defining conversation of the coming decade and, likely, century. Ensuring that we fully understand the complex framing of these issues and ask the challenging questions that will impact both policy and practice is critical if we are to harness the opportunity of AI and ensure that its impacts are justly and equitably experienced across society.

The views and opinions of third-party content providers are solely those of the author and not Fidelity Charitable. Fidelity Charitable does not guarantee the accuracy of the information provided by such third parties.