Nonprofit Current AI launches open-source infrastructure to challenge Big Tech dominance

Current AI, a nonprofit organisation established to develop open-source, multilingual artificial intelligence infrastructure, is positioning itself as a public alternative to private, market-driven AI systems. Backed by $400 million in funding from governments and philanthropic bodies, including the French government, Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce, the entity operates as a public-private partnership where funders are explicitly not investors. The organisation’s mission prioritises consent, context, and data sovereignty for underrepresented languages and cultures, contrasting sharply with the expansionist approaches of major technology firms.

The nonprofit recently launched Suno Sutra, an offline AI device supporting 22 Indian languages, developed in partnership with Bhashini, the Indian government’s AI language division. The pocket-sized device functions without internet access, allowing users in rural areas to research information in their local dialects. The device is open-sourced, enabling developer communities to build upon its framework. Current AI was founded in February 2025 by Martin Tisne, with Ayah Bdeir joining as chief executive officer in January after leading Mozilla’s AI strategy.

In a move to decentralise AI development, Current AI allocated $3.2 million in grants to four organisations across Kenya, Lebanon, and Brazil. The funding supports local data control and language preservation, including projects by Masakhane in Kenya to build datasets for over 50 African languages, the Institute for Worldmaking in Lebanon to digitise Arab cultural history, Portal sem Porteiras in Brazil to create offline AI tools for Indigenous communities, and the African Internet Rights Alliance to develop audit tools for accountability.

The organisation also launched Alpha Chat, an open-source chatbot assembled in seven weeks by a coalition of ten entities, including Hugging Face, Mozilla, and the MIT Media Lab. Each contributor provided components of the stack, such as language models and safety tooling. Additionally, Current AI has partnered with Tokyo-based startup Sakana AI to build a shared open-source AI stack designed to support the Japanese language and culture, as well as communities in the Global South that have been largely ignored by dominant AI systems.

Bdeir has drawn a distinction between Big Tech’s multilingual efforts, which she describes as market-expansion driven, and Current AI’s approach, which prioritises consent and context. She argues that while half the world’s spoken languages face extinction, English-driven models leave most cultures behind. The organisation’s vision is an open system modelled on the early web, where improvements benefit everyone, no one is locked out, and communities retain control of their own data.