Humanitarian organisations in Africa may be falling behind the race to adopt Artificial Intelligence (AI) in their operations, slowing down predictions of crises.
The new revelations came even as the technology is taken on board humanitarian work across the world. A study published on Monday says organisations are failing to keep pace with frontline workers who are already deploying the technology in crises.
The report, released by the Humanitarian Leadership Academy and Data Friendly Space, surveyed 2,539 humanitarian workers across 144 countries, many of those in Africa and in places were humanitarian crises have risen, between May and June 2025.
It found that while 93 percent of aid workers have tried AI tools, only 8 percent of humanitarian organisations have integrated them into operations, exposing what researchers describe as a “humanitarian AI paradox.”
“The humanitarian AI paradox shows us that innovation is happening despite institutional gaps, not because of them. We need coordinated sector-wide approaches to practical implementation,” said Madigan Johnson, Head of Communications at Data Friendly Space.
She noted that the Humanitarian organisations must act to bridge the AI implementation gap or risk compromising their ability to respond effectively to increasing global crises, as workers already using these tools lack proper training and governance frameworks.
Latest data shows that most people are increasingly concerned with AI and its impact on their daily lives. Google Trends data, for example, shows that searches for AI-related content in Kenya surged dramatically: around 150 percent between January–July 2024 compared to the same period in 2023, and even up to 270 percent year-on-year, with a 400 percent increase over five years.
Specific increased searches include “what is AI” (+60 percent), “how to use AI” (+90 percent), “AI jobs” (+230 percent), “AI courses” (+120percent), and breakout search terms like "AI-generated images" or “logo maker AI” spiking over 5,000 percent.
In 2024, Kenya ranked 13th globally in AI search interest, with dramatic increases in queries like “what is AI?” (+60 percent), “how to use AI” (+90 percent), and “AI jobs” (+230 percent).
At least 42 percent of Kenyan internet users have also used ChatGPT in the past month, doubling the global average of 22 percent and indicating that Kenya now leads the world in ChatGPT usage per capita.
Kenya, like peers in Uganda, Somalia, South Sudan, Ethiopia, Chad and Tanzania have been recipients of refugees fleeing crises in the region. They have also seen people displaced by one natural disaster or another. This makes humanitarian responses in these countries a daily affair.
The report did not mention specific organisations or workers interviewed for the assessment on AI. But it shows that a slow response could hinder organisational preparedness for humanitarian crises.
According to the report, in Sub-Saharan Africa, humanitarian staff are turning to AI chatbots to train caregivers in child protection. After addressing privacy concerns, uptake increased, allowing communities with limited access to trainers to gain vital knowledge.
In the Middle East, offline AI assistants are being piloted in Lebanon to strengthen operational security, ensuring aid workers can access critical support without risking sensitive data leaks.
Meanwhile, in Afghanistan, AI-driven analysis has reduced dropout rates in women’s empowerment programmes by 30 percent, offering hope in a country where girls’ access to education has been severely restricted.
And in Ukraine, humanitarian workers have developed AI-generated safety training materials for children caught in conflict — content that would otherwise take weeks to prepare being delivered in days.
Lucy Hall, a Data and Evidence Specialist at the Humanitarian Leadership Academy, said these examples demonstrate how frontline staff are leading the charge.
“It’s the individual humanitarians who are driving the AI transformation. Organizations and leaders now need to build systems shaped by those already using AI to respond more effectively to crises," she said.
Despite this creativity, the report warns that humanitarian organizations’ failure to adopt AI systematically poses serious risks. Workers are heavily reliant on commercial platforms such as ChatGPT and Claude — with 69 percent using them — rather than tools built for humanitarian settings.
This reliance raises concerns about bias, data sovereignty, and safeguarding vulnerable populations, particularly as 64 percent of workers’ report receiving little or no AI training.
"The stakes are high. Disasters have risen by 35 percent over the past decade, while humanitarian needs are at record levels, with a $50–73 billion funding gap left by donor governments in the Global North," the report stated.
"By comparison, AI adoption is far higher in other sectors — 47 percent in the private sector and 23 percent in healthcare — highlighting how far behind aid organisations remain at just 8 percent," it continued.
