UNDP Every Language Matters Open Call

The United Nations Development Programme (UNDP) is inviting people who are making AI work in their own language and culture to join Every Language Matters, its initiative for AI that serves every community, no matter the language they speak. AI systems work reliably in fewer than 100 of the world's 7,000+ languages, and ELM brings together governments, researchers, funders, technology partners and communities to change that, building on two years of UNDP's language accelerators on the ground.

What you get:

• A route into UNDP's Every Language Matters network, which convenes a global pool of experts and innovators to set shared standards for language-inclusive AI
• A way to put your work, and what is holding it back, in front of UNDP's Digital, AI and Innovation Hub
• A link to ELM's work connecting projects with partners and new resources to scale

What you'll do:

• Describe what you are building, for whom, and the languages you work in
• Name the stage where you are most stuck (data collection, data curation, model training, deployment or re-use) and what support would unlock progress
• Share one success or proven method from your work that others could learn from

Key dates:

• Applications close November 15, 2026

More information:

• UNDP's eight questions about ELM: https://www.undp.org/digital-innovation/elm-eight-questions
• The Data-to-AI Value Chain behind the initiative: https://www.undp.org/digital-innovation/elm-data-to-ai-value-chain

• Anyone working on making AI serve their language and culture, in any country
• Open to people at research or academic institutions, start-ups and private companies, civil society and community organisations, government and public institutions, and open-source or volunteer communities
• Apply through the Join Now button on UNDP's page, which opens a short online form: your work in up to 100 words, your countries and languages (with ISO codes if you know them), your focus areas such as health, education or agriculture, the language technologies you work with (speech recognition, text-to-speech, machine translation, large language models, chatbots and more), your biggest bottleneck and one success in up to 50 words each