African philanthropy leaders call for stronger oversight of AI in grantmaking
Artificial intelligence could help philanthropic organisations review applications and carry out due diligence faster. But leaders in the sector are warning against allowing the technology to determine which organisations receive funding, arguing that such decisions still require human judgment and an understanding of local contexts.
The issue was featured in discussions at the African Philanthropy Forum (APF) Conference in Cape Town, South Africa, held from October 25 to 27. Under the theme “Bridge the Divide; Reimagine the Future,” the conference explored how philanthropy can respond to technological change while addressing Africa’s development priorities.
For grantmakers, the question is not simply whether AI can make their work more efficient, but how far they should rely on it when making decisions that affect organisations and the communities they serve.
Efficiency must not replace human judgment
Toyin Akinniyi, regional portfolio director for Africa at Luminate, said AI could reduce the time spent on due diligence but cautioned against using it to automate decisions about funding.
“It will help reduce the amount of time used on due diligence, yes, but if it is used to automate people’s roles or decide who gets in or not, that will not work,” she said.
Reviewing an application involves more than processing information. Funders also need to understand an organisation’s work, its relationship with the community and the circumstances in which it operates. AI may help with the paperwork, but Akinniyi’s warning was against allowing efficiency to take precedence over the judgment involved in deciding who gets support.
Jessica Osborn, CEO of Caribou, also emphasised the human element in grantmaking, cautioning against allowing technology to determine the substance of proposals.
“The substance applicants bring should embody the human element and the trust their communities have in them. Humanism is irreplaceable. Human judgment is critical,” she said.
Her remarks speak to a concern for organisations whose work depends on relationships and trust that may not be adequately captured in an application or dataset.
Who shapes the technology?
The discussion also raised questions about Africa’s role in developing AI, not just adopting it.
Akinniyi urged funders to look beyond individual innovations and consider the research, expertise and governance needed to support the wider technology ecosystem. She pointed to the limited representation of African researchers and voices in the data sector.
“When we fund, we should think about the entire ecosystem. We don’t have a lot of researchers or African voices in the data sector,” she said.
For philanthropic organisations, that means considering whether their investments help build African expertise and influence over the technology, rather than focusing only on tools that are already available.
Janade du Plessis of Launch Africa Ventures said philanthropists could use their influence to give policymakers time to develop appropriate safeguards as AI advances.
“As a philanthropist, you can use your voice to slow the pace of AI to create space for policies and regulations,” he said.
He also called for human and machine capabilities to work together, with African influence established before AI takes on decision-making roles.
The concern extends beyond the risks of individual tools. As AI becomes more widely used, African institutions will need the expertise and influence to question how these systems work and whose interests they serve.
Preparing organisations for AI
That responsibility also falls on the organisations adopting the technology. APF board member Nyimpini Mabunda highlighted the need to make AI literacy more accessible to people in decision-making positions, many of whom do not have technical backgrounds.
Leaders need to understand what AI can do, where it falls short and what its limitations mean for their organisations. Without that knowledge, adopting a tool does not necessarily mean being equipped to assess its output.
Osborn similarly pointed to governance as a challenge, arguing that deciding how to use AI can be harder than choosing the technology itself. Organisations need staff with the right skills, clear principles and oversight to guide its use.
These questions carry particular weight in grantmaking, where decisions can determine which organisations have the resources to continue their work. Funders need to establish where AI can assist and where a decision requires closer human scrutiny.
Akinniyi also called for greater collaboration between philanthropy and other capital holders, arguing that investment in AI and technology needs a longer-term approach.
“The point of philanthropy is about derisking all projects. The way we have funded AI and tech doesn’t reflect the power we have as philanthropists,” she said, adding that the work required collaboration and patient capital.
Her argument was for funders to consider the wider needs of the sector, including research, governance and the expertise required to guide technological development, rather than concentrating on individual projects.
The discussions at the APF conference reflected a tension facing philanthropy as AI becomes more accessible. The technology offers opportunities to improve how organisations work, but its use also raises questions about accountability, African representation and who ultimately makes decisions.
For grantmakers, the challenge will be to take advantage of AI where it is useful without handing over decisions that depend on human judgment and an understanding of the communities their funding is meant to serve.
