Current AI, the nonprofit backed by $400 million in committed funding from the French government, Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce, deployed $3.2 million in grants to four organizations and launched an open-source chatbot at the AI for Good Summit in Geneva, TechCrunch reported on July 19.
The organization also unveiled Suno Sutra (“listening chronicles” in Hindi), a pocket-sized offline device built with Bhashini, the Indian government’s AI language division. The device runs AI assistants in 22 Indian languages without requiring internet connectivity. It is open-sourced for developer communities to build on.
The Agent Infrastructure Angle
Most agentic frameworks today assume cloud connectivity and API access. Composio, CrewAI, LangChain, and the major agent orchestration platforms all require persistent internet connections and API calls to function. Suno Sutra demonstrates edge-first, locally-run AI infrastructure that operates independently of cloud services.
For the agent ecosystem, the question is whether offline-first models can support the kind of multi-step reasoning and tool use that defines agentic workflows. The device currently runs assistive AI rather than autonomous agents, but the underlying architecture of local model inference on resource-constrained hardware is relevant to any team building agent deployments outside the hyperscaler cloud.
Where the Grants Went
Current AI’s first cohort of $3.2 million went to four organizations, according to TechCrunch:
Masakhane in Kenya is building AI datasets across more than 50 African languages for health, farming, and education. Lebanon’s Institute for Worldmaking is digitizing Arab cultural history into machine-readable databases controlled by communities rather than tech companies. Brazil’s Portal sem Porteiras is building offline AI tools with Indigenous Amazon communities, keeping data within the territory. Kenya’s African Internet Rights Alliance is developing audit tools to hold AI systems accountable across the continent.
The Public-Private Model
Current AI CEO Ayah Bdeir, who joined in January after leading Mozilla’s AI strategy, described the organization as a “public-private partnership” to TechCrunch. The French government seeded it with $100 million. Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce brought the total to $400 million. “They’re not investors; they’re funders,” Bdeir told TechCrunch.
The nonprofit also launched AlphaChat, an open-source chatbot assembled in seven weeks by a coalition of ten organizations including Hugging Face, Mozilla, and MIT Media Lab. Each contributor brought a piece of the stack: language model, safety tooling, and computing power.
A separate deal with Tokyo-based Sakana AI aims to build a shared open-source AI stack supporting Japanese language and culture, plus communities across the Global South.
Why It Matters for Agent Builders
Current AI’s model represents an alternative to the VC-funded agent infrastructure stack. Where venture-backed startups optimize for enterprise seats and usage-based revenue, Current AI optimizes for reach without revenue. Half the world’s spoken languages face extinction, according to UN data cited by TechCrunch. English-dominant training data means AI systems encode one linguistic worldview while discarding others.
Whether $400 million in philanthropic funding can sustain open infrastructure at the scale needed to compete with proprietary systems is an open question. But for agent builders, the offline-first pattern and community-controlled data governance model offer architectural alternatives worth tracking as agent deployments expand beyond English-speaking enterprise markets.