Current AI Unleashes Initiative to Create Global Open Platform for Artificial Intelligence

A farmer living in rural India captures a photograph of a plant that appears to be dying. She wishes to research this issue online, but she does not speak English. Fortunately, there is no need for her to rely on language barriers to access valuable information.

This kind of challenge is exactly what nonprofit Current AI aims to address by developing open, publicly accessible AI infrastructure. At the India AI Summit held in February, the organization partnered with Bhashini, the Indian government’s unit responsible for advancing AI language technologies. Together they created Suno Sutra, a compact offline device that runs AI functions in 22 different Indian languages without requiring an internet connection. The name “Suno Sutra” translates to “listening chronicles” in Hindi. Ayah Bdeir, Current AI’s CEO, explained in an interview with TechCrunch that India boasts hundreds of distinct languages and dialects, yet current AI systems fail to represent them adequately. The device is open-source, allowing developers from around the world to build upon its foundation.

Founded in February 2025 by Martin Tisne, Current AI has been advancing its efforts rapidly. Last month the organization allocated $3.2 million in grants to several projects across four different organizations. Most recently, it introduced an open-source AI chatbot at the AI for Good Summit in Geneva.

Bdeir joined Current AI in January after serving as head of Mozilla’s AI strategy. Prior to that, she was the founder of littleBits, a STEM education company that reached millions of children before being acquired by Sphero in 2019.

According to Bdeir, Current AI operates as a public-private partnership that brings together governments, private companies, and philanthropic organizations to fund technologies that serve the public good. The French government initially provided $100 million to support the organization, with additional funding coming from the Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce. This brings the total committed funding to $400 million. Bdeir emphasized that these entities function as funders rather than traditional investors.

The core issue Current AI seeks to solve is straightforward: almost all major AI systems today, including those developed by OpenAI, Google, and Anthropic, are owned by private companies. Bdeir stated that if AI is truly meant to be a transformative technology that influences every aspect of people’s lives, there must exist a public alternative — something as accessible to everyone as the World Wide Web and available at no cost.

Half of the world’s spoken languages are currently at risk of going extinct. Bdeir added that since English drives the development of the largest language models and AI systems, many other languages, along with their associated cultures and communities, remain marginalized.

More than just language

The ability of an AI to speak a particular language is only one aspect of what it needs to learn. Bdeir explained that language serves as the medium through which knowledge, traditions, memories, and identity are passed down from one generation to the next. Therefore, if a technology cannot communicate in a community’s native language, it also cannot preserve that community’s culture.

Her vision for Current AI is an open system modeled after the early versions of the web, where improvements benefit all users, no one is excluded from participation, and communities retain control over their own data.

The first round of grants issued by Current AI last month allocated $3.2 million to four organizations located in Kenya, Lebanon, and the Brazilian Amazon.

In Masakhane, Kenya, the project involves creating AI datasets in more than 50 African languages for use in healthcare, agriculture, and education. The Institute for Worldmaking in Lebanon is working to digitize Arab cultural history and contemporary practices into machine-readable databases that are controlled by local communities rather than tech companies. Portal sem Porteiras in Brazil is developing offline AI tools alongside Indigenous Amazon communities, ensuring that data remains within those regions. Lastly, the African Internet Rights Alliance in Kenya is creating auditing tools to hold AI systems accountable across the continent.

Who controls the data?

When addressing the issue of data ownership, Bdeir was direct. She told TechCrunch that while there are various models and proposals regarding who should own data within different communities, one thing is clear: it should not be controlled by a Silicon Valley company aiming to enrich just a small number of people.

The nonprofit’s approach involves storing models and data locally, involving community experts at every stage of development, and incorporating consent protocols into the process so that communities can stop any initiative at any time.

None of Current AI’s funded projects have yet fully resolved this issue. However, Bdeir sees this as exactly what is intended. She said each project has incorporated this question into its work, rather than accepting the common practice of using complexity as an excuse to let governments or tech companies make decisions on behalf of everyone else.

Regarding how much progress can be achieved with a $3.2 million budget distributed across four organizations, Bdeir emphasized that scale is not always the best measure — this reflects the thinking typical of large technology companies. She illustrated potential outcomes by describing an Indigenous elder in the Brazilian Amazon using a tool developed in Kenya to share ecological knowledge in their own language.

Developing the technical foundation

Earlier this month in Geneva, Current AI launched Alpha Chat, an open-source chatbot that was created within seven weeks through collaboration among ten organizations, including Hugging Face, Mozilla, and MIT Media Lab. Each participant contributed a component necessary for the system to function, such as a language model, safety tools, or computing resources.

Current AI has also entered into a partnership with Sakana AI, a Tokyo-based startup renowned for its work on what it calls Sovereign AI. The two organizations plan to develop a shared open-source AI stack designed not only to support the Japanese language and culture but also to assist communities in the Global South that have been largely overlooked by mainstream AI systems.