The authors identified four archetypes describing philanthropic funders’ approach to AI: The Curious, The Doers, The Dreamers, The Skeptics. The authors did not find major differences found across foundations based on their geographic location, though there were differences found based on their mission and values. Feminist and social justice funders in the Global South demonstrated more skepticism than others. Even in seemingly benign or straightforward applications, AI systems demonstrate unpredictable behavior, leading to most foundations proceeding with caution. The rapid pace of AI has meant that formal AI policies are rare, though foundations align on some common “red lines” that guide internal use of AI.
Higdon et al. (Mon,) studied this question.