Gates Foundation pledges $1 billion for AI access as Bill Gates warns governments lag on risks

A two-year pledge for global access

The Gates Foundation announced in its annual Goalkeepers report that it will spend at least $1 billion over the next two years to expand access to artificial intelligence. The funding forms part of the foundation's planned $9 billion annual budget, directing resources toward education, healthcare, and agriculture in low-income regions. According to the foundation, 40% of the funds will target healthcare initiatives, including drug discovery and diagnostic support for frontline clinical staff. Another 40% is dedicated to education tools, while agriculture and digital foundations will each receive 10%. Early deployments cited in the report include clinic diagnostic software in Kenya, educational platforms in Sierra Leone, and a farming advice service supporting 740,000 farmers in India.

Linguistic disparities in training data

A core component of the initiative involves building dataset foundations to address linguistic imbalances in current models. Early large language models drew more than 90% of their training data from English sources, resulting in low error rates for English speech recognition of under 6%. In contrast, recognition error rates for West African languages such as Yoruba are ten times higher. Web content distribution reflects similar imbalances, with English representing roughly 49% of available text online. Spanish accounts for 6%, Japanese for 5%, and Chinese for 1.3%, prompting China's National Data Administration to construct validated training datasets to expand domestic capabilities.

Calls for state intervention and regulation

Alongside the funding announcement, Microsoft co-founder Bill Gates criticized the preparedness of public authorities, asserting that public institutions lag behind the pace of technological development. Gates stated that private industry cannot be expected to govern itself and argued that government regulation is necessary to manage societal risks.

I don't think any government is nearly as deep on this as they have to be.

In recent months, Gates shifted from earlier optimism to raise concerns regarding labor market disruption, cybersecurity attacks, automated disinformation, and dependency on AI companions. In August, he proposed designating Human Reserved jobs to protect roughly 40% of occupations from automation, paired with tax mechanisms on AI tokens and robotic systems. Foundation leadership also addressed foreign aid policy, with Chief Executive Mark Suzman affirming continued engagement with American lawmakers and Donald Trump's administration despite foreign assistance cuts.

We haven't given up on the U.S. We will acknowledge it's been a challenging period with the U.S. government on this. But we believe there is still a case to be made.

Geopolitical friction and industry divisions

Gates engaged with international officials regarding AI governance, speaking with Trump's team in the United States and scheduling talks with Chinese President Xi Jinping later in the year. He compared the necessity of cross-border alignment to disaster cinema where rival nations unite against external threats.

AI is kind of like this alien intelligence. It's here, and we better do like it shows in those movies.

However, international approaches remain divided. Trump dismissed arguments for slowing AI development and told Nvidia chief Jensen Huang that fear regarding the technology is a hoax. Meanwhile, leaders across tech companies, including Anthropic's chief executive, OpenAI chief Sam Altman, SpaceXAI founder Elon Musk, and Google DeepMind chair Demis Hassabis, supported calls for a measured pace following safety resignations within the sector.