New York: Organisations from across the global AI ecosystem today announced a joint commitment to help close one of AI’s most significant access gaps: providing safe, effective AI services in the languages people speak around the world.
A group of 60 signatories are uniting behind a shared five-year goal: for an estimated 3.4 billion people who speak languages currently underrepresented in today’s AI models to be able to use AI tools in their own language and voice, African Media Agency (AMA) reported.
Initial signatories reflect the range of expertise needed to make progress, from locally led research and implementation to AI model development, funding, and public policy. They include: frontier AI labs, government ministries and departments, philanthropies, developers, and implementing organisations from across the globe. The joint commitment invites others across the ecosystem to join the effort and contribute their own resources and expertise.
Today, only a small percentage of the world’s roughly 7,000 languages are considered well-resourced enough to support strong AI capabilities. Many others – often described as “low-resource languages” – are underrepresented in the data, tools, and benchmarks used to build and test AI. As a result, people who speak them, including many in communities that are already less well served by today’s technologies, can be underserved by systems that are less accurate, less useful, or less able to understand how they communicate.
The challenge goes beyond translating text: voice can be particularly important where typing or text-based interfaces are less practical or accessible, allowing people to interact with AI through natural speech. Poor language performance is not just an inconvenience. Dialect, slang, idioms, and cultural context can change meaning, with potentially serious consequences when accuracy matters most.
Closing this gap could make AI-enabled information and services more accessible and useful to billions of people, across a range of sectors including health, education, agriculture, financial services, or public services.
Participating organisations, each contributing their individual resources and expertise, will work together to accelerate progress across four areas:
- Building the open language layer: the shared, safe data infrastructure that every builder can draw on using open licenses
- Tracking progress honestly: assessments and benchmarks that measure real gains against the global goal
- Turning language data into working tools: models and applications usable by any AI builder, not just those with the most resources
- Reaching people safely: guided throughout by responsible practices that protect privacy, consent, and data sovereignty
