Nonprofit Current AI races to build the World Wide Web of AI, free for all

Summary

Current AI, a non-profit organization, is backed by $400 million in public and philanthropic funding. The organization is building open-source AI infrastructure. This infrastructure includes an offline device supporting 22 Indian languages and a chatbot developed by a coalition. Current AI positions itself as a public-interest alternative to AI systems under corporate control.

Why it matters

This is one of the most structural and serious initiatives to treat AI as public infrastructure rather than a proprietary product. The venture directly challenges the assumption that meaningful AI solutions can only be provided by major tech companies. Its focus on language preservation, data sovereignty, and community consent for marginalized groups fills a real and growing gap that commercial multilingual models have largely ignored.

Highlights

• Suno Sutra: An open-source, offline mobile device developed with India's Bhashini platform, running AI in 22 Indian languages
• $400 million in committed funding from the French government, Ford/MacArthur Foundations, DeepMind, and Salesforce; institutions are explicitly positioned as funders rather than investors
• A $3.2 million grant round distributed to Kenya, Lebanon, and the Amazon region of Brazil for community-controlled AI datasets and tools
• AlphaChat: An open-source chatbot developed in seven weeks by a coalition of 10 organizations, including Hugging Face, Mozilla, and the MIT Media Lab

Critical take

The article reads more like a press release than independent journalism. Every concrete claim is attributed to CEO Bdeir. There is no third-party benchmarking of Suno Sutra's actual language performance. The emphasis on the chatbot being developed in seven weeks deliberately obscures the fact that the underlying models, compute infrastructure, and security tools came from well-funded institutions. This makes the open-source narrative somewhat boastful. The $3.2 million distributed across four different regions raises the legitimate question: is this meaningful infrastructure, or just well-publicized pilot projects?