Nonprofit Current AI builds open, free AI for underserved languages

Current AI, a nonprofit founded in February 2025 by Martin Tisne, is positioning itself as a counterweight to the private ownership of AI. The organization is led by Ayah Bdeir, who joined in January after leading Mozilla's AI strategy and previously founded littleBits, an education company that reached millions of children before being sold to Sphero in 2019. Bdeir articulated the nonprofit's core mission in an interview: "If AI is truly a transformative technology, if it's going to change every aspect of everyone's life, there has to be a public alternative. Like the World Wide Web, available to anyone, for free."

The funding structure reflects this ambition. Current AI has secured $400 million（約640億円） in total committed funding, including a $100 million（約160億円） seed from the French government alongside commitments from the Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce. Bdeir was explicit that these organizations are "funders, not investors," and that Current AI operates as a public-private partnership bringing together governments, companies, and philanthropies to support public interest technology. Last month, the nonprofit allocated $3.2 million（約5.1億円） in grants to projects across Kenya, Lebanon, and the Brazilian Amazon. Masakhane in Kenya is building AI datasets across more than 50 African languages for health, farming, and education. Lebanon's Institute for Worldmaking is digitizing Arab cultural history and contemporary practice into machine-readable databases that communities—not tech companies—control. Brazil's Portal sem Porteiras is developing offline AI tools with Indigenous Amazon communities, ensuring data remains within the territory. Kenya's African Internet Rights Alliance is creating audit tools to hold AI systems accountable across the continent.

One of Current AI's earliest concrete products is Suno Sutra (Hindi for "listening chronicles"), a pocket-sized offline device developed in partnership with Bhashini, the Indian government's AI language division. Announced in February at the India AI Summit, Suno Sutra runs AI in 22 Indian languages without requiring an internet connection and is open-sourced for developer communities. Bdeir emphasized the problem it addresses: "In India, there are hundreds of different languages and dialects, and right now AI is not representing them."

Bdeir drew a sharp distinction between Current AI's approach and Big Tech's multilingual expansion. "Big tech builds multilingual models to expand their market, regardless of consent or context," she said. She cited a concrete example: "For Indigenous languages, missionary Bible translations become training data before communities have set any rules." For Bdeir, the issue goes beyond language. "Language is how knowledge, tradition, memory and identity get carried from one generation to the next. So when a technology can't speak your language, it can't hold your culture either," she said. On data ownership, she was direct: "There are different models and proposals for who owns data in various communities, but one thing is sure: it shouldn't be a company in Silicon Valley trying to make a select few thousand people wealthier."

Current AI's immediate technical push includes Alpha Chat, an open-source chatbot launched this month in Geneva at the AI for Good Summit. The tool was assembled in seven weeks by a coalition of ten organizations, including Hugging Face, Mozilla, and MIT Media Lab, each contributing a piece of the stack—a language model, safety tooling, and computing power. The nonprofit has also struck a deal with Sakana AI, a Tokyo-based startup known for its work on what it calls Sovereign AI. The two organizations plan to build a shared open-source AI stack designed to support Japanese language and culture, as well as communities across the Global South that dominant AI systems have largely ignored. When asked about progress given the $3.2 million（約5.1億円） budget split across four organizations, Bdeir rejected scale as the primary measure. "Scale is not always the measure. That is the Big Tech paradigm," she said. "This could look like an Indigenous elder in the Brazilian Amazon using a tool built in Kenya to be able to pass down ecological knowledge in their own language."

Current AI emerges at a moment when the dominance of English-language AI models has created a cultural and linguistic divide. Half the world's spoken languages face extinction, and as CEO Bdeir noted, English-driven large language models mean that entire cultures and communities are left behind. The nonprofit's founding in February 2025 by Martin Tisne, with Bdeir joining in January after leading Mozilla's AI strategy, reflects a deliberate pivot: where Big Tech builds multilingual systems primarily to expand markets, Current AI is structured explicitly to keep control in the hands of communities and local governments.

The organization's approach differs from typical commercial AI development in both governance and data ownership. Rather than centralizing training data and models in corporate servers, Current AI emphasizes local deployment, community consent, and explicit data ownership protocols. The $3.2 million（約5.1億円） grant cohort announced last month—distributed across Masakhane (Kenya), the Institute for Worldmaking (Lebanon), Portal sem Porteiras (Brazil), and the African Internet Rights Alliance (Kenya)—shows this principle in practice: building health and farming datasets in African languages, digitizing Arab cultural history under community control, and creating offline tools for Indigenous Amazon communities. Bdeir acknowledged that none of these grantees has fully solved the data ownership question yet, but framed that ongoing work as the point itself—embedding consent and control into the pipeline rather than treating complexity as an excuse for corporations to decide on behalf of communities.