MANY MINDEDA $400M nonprofit bid to make AI a public utility, not a product
Current AI, a nonprofit building open, public AI infrastructure, was founded in February 2025 by Martin Tisné. Its CEO, Ayah Bdeir, joined in January after leading Mozilla's AI strategy and, earlier, founding littleBits, which sold to Sphero in 2019. The French government seeded the effort with $100 million; the Ford Foundation, MacArthur Foundation, DeepMind and Salesforce have since brought total committed funding to $400 million.
The money is already moving into tools for languages the big labs neglect. Last month Current AI split $3.2 million in grants among four groups in Kenya, Lebanon and the Brazilian Amazon. In February it worked with Bhashini, India's government language-AI division, on Suno Sutra, an offline pocket device running AI in 22 Indian languages and released as open source. It also shipped Alpha Chat, an open source chatbot built in seven weeks by ten organizations including Hugging Face, Mozilla and MIT Media Lab, and signed a deal with Tokyo's Sakana AI for a shared open stack.
The premise, in Bdeir's framing, is that every major AI system today — OpenAI, Google, Anthropic — belongs to a private company, and a free public alternative would keep this tool a common good. With the article noting that half the world's spoken languages face extinction, coverage of the long tail is itself a form of access.
Keep the scale honest. This is early-stage work, and $3.2 million split four ways is small against the problem. None of the grantees have settled who owns community-generated data, and much of the vision comes from the CEO's own account. Watch whether the $400 million commitment converts into infrastructure others can actually build on.
Source: TechCrunch