Large cuts to international aid in 2025 reshaped development funding and put health and humanitarian services at risk across many low- and middle-income countries. The shift began on 20 January when the US ordered a suspension of almost all foreign aid contracts for 90 days pending a review. That suspension triggered mass layoffs, an immediate halt to US-funded services worldwide and ultimately the closure of the US Agency for International Development (USAID), an institution active since 1961; the US also withdrew from the World Health Organization.
Other donors followed with reductions: the UK cut its overseas aid to free funds for defence, and France, Germany and the Netherlands made significant cuts. Analysts warned the combined effect left 26 low- and middle-income countries, with nearly 1.4 billion people, highly vulnerable. Investigations reported broken health systems, stopped water and sanitation projects in northern Idlib, and about 200 humanitarian organisations pausing operations. In Africa, patients in Nairobi’s Mathare settlement faced HIV drug stock-outs, community programmes in Uganda were on the brink, malaria insecticide campaigns stopped in several countries, and malaria gene research was put at risk.
A 2025 UNAIDS report said 1.3 million people were newly infected with HIV in 2024 and warned that the US funding withdrawal could lead to an additional 6 million infections and 4 million HIV-related deaths by 2029. The US had invested over US$110 billion in the global HIV/AIDS response through PEPFAR since 2003 and spent US$12.4 billion in 2024. Responses included philanthropic pledges and plans for greater self-reliance; Bill Gates pledged most of his foundation’s wealth to Africa and African leaders proposed meeting 60% of vaccine needs domestically by 2040. Despite the crisis, there were advances such as an affordable twice-yearly HIV drug, a promising malaria treatment and a single-dose dengue vaccine made in Brazil. COP30 in Brazil agreed measures to mobilise US$1.3 trillion annually by 2035 and to increase adaptation finance by 2025 and 2035. The year also saw local resilience efforts and growing uses of AI in health.
Difficult words
- suspension — temporary stopping of an activity or service
- vulnerable — likely to be harmed or affected badly
- stock-out — periods when a necessary supply runs outstock-outs
- philanthropic — giving money or help for public good
- self-reliance — ability to meet needs without outside help
- withdrawal — act of removing support or funding
- resilience — ability to recover from shocks or problems
- adaptation — changes to cope with new conditions
Tip: hover, focus or tap highlighted words in the article to see quick definitions while you read or listen.
Discussion questions
- How might cuts to international aid affect health and humanitarian services in low- and middle-income countries? Give examples from the article.
- What are the possible benefits and limits of greater self-reliance and philanthropic pledges mentioned in the text?
- In what ways could local resilience efforts and growing uses of AI help health systems during funding shortages?
Related articles
Latin American groups build AI to study gender violence
Groups in Latin America create open, local AI tools to study gender inequalities and violence. Projects like AymurAI search court documents, protect sensitive data on local servers and help governments and civil society with evidence.
Crop losses threaten food security across Africa
Researchers say climate shocks, pests and diseases are increasing crop losses across Africa. Erratic rainfall, flooding and biological threats reduce yields, lower incomes and affect national food supplies and global commodity markets.
Australia's harsh summer: fires, floods and heat
Australia experienced a run of heatwaves, bushfires, storms and floods across the January–February summer, linked to the worsening climate crisis. The events damaged ecosystems, affected people and added pressure on services and insurance.
New training method helps models do long multiplication
Researchers studied why modern language models fail at long multiplication and compared standard fine-tuning with an Implicit Chain of Thought (ICoT) method. ICoT models learned to store intermediate results and reached perfect accuracy.