The Gates Foundation warned in its annual Goalkeepers Report published on Tuesday, Sept. 15, 2026, that artificial intelligence (AI) could widen global inequality unless the technology is specifically developed to include data from lower-income nations. The report stated that billions of people risk being excluded from the benefits of AI because current models often do not understand their languages or local living conditions.
The foundation noted that the disparity is driven by the data used to train large language models (LLMs). According to the report, more than 90% of the data used for early LLMs came from English-language sources. While leading AI speech-recognition systems have an error rate of less than 6% in English, the error rate is 10 times higher for Yoruba, a language spoken widely in Nigeria and West Africa.
The report also detailed how AI tools designed for wealthy nations may fail when applied elsewhere without local context. For example, a system built for commercial farmers in Iowa may not accurately account for the specific crops, pests, or market constraints faced by smallholder farmers in Ethiopia. To address these gaps, the Gates Foundation announced it is committing at least $1 billion over the next two years to support digital infrastructure, local-language datasets, and AI applications in healthcare, education, and agriculture.
The foundation warns that without deliberate intervention, market forces will prioritize high-paying institutions in wealthy countries. Bill Gates stated in the report that decisions made in the next 12 to 18 months regarding how AI is built and funded will determine whether the technology reaches those who have the least. If accessible, AI could help address shortages of doctors and teachers by providing specialist knowledge through mobile devices.
Early projects highlighted in the report include an AI system that has improved diagnostic accuracy in Kenyan clinics, education tools showing learning gains in Sierra Leone and the U.S., and an agricultural service currently used by 740,000 farmers in India. Foundation officials will monitor whether these investments can scale such projects before market standards become fixed.