Current AI, a nonprofit founded in 2025, is attempting to build what its leaders describe as a public alternative to today's largely corporate-controlled artificial intelligence ecosystem.

Rather than developing another proprietary AI model, the organization is investing in open infrastructure, multilingual models, community-owned datasets and freely available software intended to function as a public digital resource similar to the early World Wide Web.

The initiative reflects a growing debate over who should control the next generation of artificial intelligence.

Most frontier AI systems are owned by a small number of private technology companies that determine how models are trained, what data they use, how they are deployed and who can access them.

Current AI argues that if artificial intelligence becomes as fundamental as electricity or the internet, public-interest infrastructure should exist alongside commercial offerings rather than leaving the technology entirely under private ownership.

The organization operates as a public-private partnership supported by governments, philanthropic institutions and technology companies.

France committed one hundred million dollars to launch the initiative, while additional backing from organizations including the Ford Foundation, the MacArthur Foundation, Google DeepMind and Salesforce has increased pledged funding to approximately four hundred million dollars.

Unlike venture-backed technology startups, the nonprofit says its supporters provide grants rather than seeking financial returns.

Its work is centered on populations that have largely been excluded from mainstream AI development.

One of its earliest collaborations produced Suno Sutra, an open-source handheld device created with India's Bhashini language initiative.

The battery-powered system can operate without an internet connection while providing artificial intelligence capabilities across twenty-two Indian languages.

The project is designed for communities where broadband access is limited and where many residents speak languages rarely represented in leading AI models.

Current AI argues that language inclusion involves far more than translation.

Languages carry historical knowledge, cultural traditions and social identity.

When a language is poorly represented in AI training data, the resulting systems often struggle to understand local customs, agricultural practices, medical terminology and community-specific knowledge.

The nonprofit contends that preserving linguistic diversity is therefore inseparable from preserving cultural diversity.

The organization also challenges the prevailing model of data ownership.

Its leadership argues that communities contributing cultural materials, oral histories or indigenous knowledge should retain authority over how those resources are collected, stored and used.

Rather than transferring datasets to centralized technology companies, Current AI promotes local storage, community consent procedures and governance structures allowing participants to withdraw data if agreed conditions are no longer met.

This philosophy extends to its grant-making program.

Earlier this year the nonprofit awarded approximately three point two million dollars to projects spanning Kenya, Lebanon and Brazil.

In Kenya, researchers are creating datasets covering more than fifty African languages to improve healthcare, education and agricultural AI applications.

Lebanese partners are digitizing Arab cultural archives under community governance, while projects in the Brazilian Amazon are developing offline AI tools alongside Indigenous communities with data remaining within local territories.

Another African initiative is building auditing tools intended to evaluate whether AI systems operate fairly and transparently.

The nonprofit has also expanded into foundational software.

At the AI for Good Summit in Geneva, it introduced Alpha Chat, an open-source chatbot assembled in only seven weeks through collaboration among organizations including Hugging Face, Mozilla and the Massachusetts Institute of Technology Media Lab.

Each participant contributed a different component, including language models, safety systems or computing infrastructure, demonstrating an alternative model of collaborative AI development.

Current AI has additionally partnered with Japanese startup Sakana AI to develop open-source infrastructure supporting Japanese while remaining adaptable for communities across the Global South.

The collaboration reflects a broader trend toward so-called sovereign AI, in which countries and regions seek greater control over language models, datasets and computing resources instead of depending exclusively on foreign technology providers.

The nonprofit's vision differs fundamentally from that of commercial AI developers.

Technology companies generally pursue multilingual capabilities to expand products into additional markets, whereas Current AI argues that communities should determine how their languages, cultural materials and traditional knowledge enter AI systems.

The organization warns that historical texts, missionary translations or publicly available documents can become training data without meaningful consultation from the communities represented.

The initiative also highlights an emerging policy debate.

Governments increasingly view artificial intelligence infrastructure as strategic national capacity comparable to telecommunications networks or energy systems.

Public investment in open models, shared datasets and locally governed computing resources could reduce dependence on a handful of global providers while giving smaller countries greater technological autonomy.

Current AI remains in its early stages, and its long-term influence will depend on whether it can translate pilot projects into durable public infrastructure adopted beyond the communities that initially helped build it.

What is confirmed is that the organization has established one of the largest publicly funded efforts to develop open artificial intelligence infrastructure, with active projects already supporting multilingual computing, community-controlled data and open-source AI technologies across multiple continents.