The Gates Foundation has warned that artificial intelligence (AI) could widen the gap between rich and poor communities unless deliberate measures are taken to make the technology accessible and relevant to those most in need.
The warning forms part of the Foundation’s 10th annual Goalkeepers Report, which calls for urgent action to ensure that AI becomes a tool for reducing inequality rather than reinforcing existing economic and social divides.
The 2026 report, titled “Make This Matter: AI, Equity, and the Choice We Can’t Delay,” argues that the future impact of AI is not predetermined and that decisions being made now will influence who benefits from the technology and how quickly.
To support efforts to close the emerging gaps, the Gates Foundation has announced plans to spend at least US$1 billion over the next two years on expanding access to AI-enabled solutions, opportunities and knowledge.
Bill Gates, Chairman of the Gates Foundation, said the challenge reflected a longstanding problem in innovation, where people with the greatest needs often had the least influence over the direction of investment and technological development.
“Much of our work has been about closing that gap. AI presents the same challenge, only at a much greater speed,” he said.
He said relying solely on market forces could result in the most advanced AI tools being developed primarily for people and institutions that could afford them, rather than for communities that could derive the greatest benefit.
The report identified language, local context, skills and affordability as key areas requiring attention if AI was to benefit a wider section of the global population.
It noted that more than 90 per cent of data used to train early large language models came from English-language sources, leaving many communities poorly represented in the systems they may increasingly depend on.
The Foundation therefore called for AI tools that could work effectively in the languages spoken by people around the world and reflect the knowledge and circumstances of the communities they were designed to serve.
It also stressed the need for countries and communities to have a role in determining how their data was collected, managed and protected.
The report further called for greater investment in people, saying doctors, farmers, teachers and developers needed opportunities to understand, evaluate and adapt AI tools to their own circumstances.
The Foundation said affordable access would also require cooperation between technology companies, governments and philanthropic organisations.
Under its US$1 billion commitment, about 40 per cent of the funding would support education initiatives, including AI tutoring and teaching tools aimed at individualising learning in the United States and other countries.
Another 40 per cent would support healthcare applications, including diagnostics, clinical decision-making tools for frontline health workers, maternal and newborn care, and the development of drugs and vaccines.
Agriculture would receive 10 per cent, with a focus on AI-generated advice for smallholder farmers based on factors such as soil, weather and crop conditions.
The remaining 10 per cent would support the digital infrastructure required for equitable AI, including datasets in languages that existing AI systems do not adequately understand.
The Foundation said its funding would help scale up the work of partners already using AI in clinics, classrooms and on small farms.
It also urged governments, technology companies, philanthropies and other stakeholders to contribute resources and expertise towards making AI more useful, accessible and affordable.
The Foundation said AI could be harnessed for good, but that this would require deliberate action and a shift in how success was measured—from focusing mainly on what AI could deliver for profitable users to considering what it could do for people who stood to benefit most.
