Nonprofit organization Current AI is accelerating efforts to create open artificial intelligence infrastructure that can be accessed, developed, and governed by communities around the world.
The organization is positioning its work as a public alternative to AI systems controlled by major technology companies. Its broader goal is to make artificial intelligence more accessible while allowing communities to retain authority over their languages, cultural knowledge, and data.
Current AI was established in February 2025 and operates through partnerships involving governments, technology companies, foundations, researchers, and nonprofit groups. The French government initially supported the organization with $100 million. Additional backing from organizations including the Ford Foundation, the MacArthur Foundation, Google DeepMind, and Salesforce has brought its total committed funding to approximately $400 million.
One of Current AI's main priorities is expanding access to AI tools for people whose languages are poorly represented in existing systems.
In India, the nonprofit partnered with Bhashini, the Indian government's language technology initiative, to develop Suno Sutra. The portable device can operate without an internet connection and supports 22 Indian languages.
The project is designed for situations where reliable connectivity and English-language services are unavailable. A farmer, healthcare worker, student, or local official could potentially use the device to access information in a familiar language without sending data to a remote commercial platform.
Current AI has made the technology open source, allowing developers and community organizations to study, modify, and expand it.
The initiative reflects a wider concern about the concentration of AI development around English and a limited number of widely spoken languages. When digital systems fail to support a language, communities can lose access to online services and may struggle to preserve cultural knowledge in emerging technologies.
Current AI is also focusing on how information is collected, stored, and used.
Rather than transferring community data to technology companies, the nonprofit supports local storage, community participation, and consent procedures. These methods are intended to give people greater influence over how their languages, traditions, and knowledge are incorporated into AI systems.
The organization recently distributed $3.2 million in grants to four groups working in Africa, the Middle East, and South America.
In Kenya, language technology group Masakhane is developing datasets covering more than 50 African languages. The work is expected to support applications in healthcare, agriculture, and education.
Lebanon's Institute for Worldmaking is converting Arab cultural history and contemporary knowledge into formats that AI systems can process. The project is structured so that participating communities, rather than outside companies, maintain control over the resulting databases.
In the Brazilian Amazon, Portal sem Porteiras is developing offline AI tools with Indigenous communities. Data created through the project is intended to remain within the participating territories.
Another Kenyan organization, the African Internet Rights Alliance, is creating tools that can evaluate AI systems and identify potential accountability, fairness, and human rights concerns.
Current AI's leaders compare their mission to the early development of the World Wide Web.
The organization wants to create shared technical infrastructure that developers can improve collectively. Tools, models, datasets, and safety systems could be made available to the public instead of being restricted to customers of individual companies.
This approach differs from the commercial model used by many leading AI developers. Private companies usually determine how their models are trained, who can use them, what information they support, and how much access costs.
Current AI argues that a technology expected to influence education, employment, healthcare, government, and communication should also have a public option.
The nonprofit is not attempting to build every part of the AI ecosystem independently. Instead, it is assembling coalitions in which different organizations contribute models, computing resources, research, safety tools, and technical expertise.
At the AI for Good Summit in Geneva, Current AI introduced Alpha Chat, an open-source chatbot developed by a coalition of 10 organizations.
The system was assembled in seven weeks with contributions from groups including Mozilla, Hugging Face, and the MIT Media Lab. Each participant provided a different part of the technology required to operate the chatbot.
The project serves as an early demonstration of Current AI's collaborative model. Rather than relying on a single company to control the entire system, the organization brings together multiple contributors to build shared infrastructure.
Current AI has also formed a partnership with Tokyo-based Sakana AI. The organizations plan to develop an open-source AI stack that supports Japanese language and culture while also serving communities in the Global South that remain underrepresented in mainstream AI systems.
Current AI faces significant challenges as it competes with technology companies that have access to vast amounts of capital, computing power, and data.
Its leaders, however, are not measuring success only through user numbers, model size, or commercial revenue. They are focusing on whether communities can use AI on their own terms and preserve knowledge that might otherwise be excluded from digital systems.
The organization's work could result in smaller, locally managed tools rather than a single global platform. A system developed for one region could also be adapted by communities elsewhere, provided that local control and cultural context remain central to the process.
Current AI's strategy represents a growing effort to treat artificial intelligence as public infrastructure. Whether the nonprofit can achieve its global ambitions remains uncertain, but its projects are testing an alternative future in which access to AI is not determined entirely by the world's largest technology companies.
Share your thoughts about this article.
Be the first to post a comment!
