A new nonprofit called Current AI is racing to build what it describes as the "World Wide Web of AI" — a free, decentralized, and globally accessible infrastructure for artificial intelligence that no single corporation or government would control.
Founded by a group of prominent AI researchers and backed by philanthropic funding, Current AI aims to stitch together open-source models, publicly available datasets, and distributed compute resources into a unified layer that anyone in the world can access and build upon.
Why this matters: The AI landscape today is dominated by a handful of Silicon Valley giants — OpenAI, Google, Meta, Anthropic — that control the most capable models, the data pipelines they're trained on, and the massive compute clusters needed to run them. Critics warn this concentration of power creates a digital divide where wealthy nations and corporations set the rules for everyone else.
"We're building the infrastructure that should have existed from the start," the organization said. "AI is too important to be locked inside corporate walled gardens. It should be a public utility — like the web itself."
Current AI's approach is multi-pronged:
- Open model hub: A repository of state-of-the-art open-source models, from small on-device LLMs to frontier-scale systems, all freely downloadable and auditable. - Federated compute network: A distributed grid of compute resources — donated by universities, data centers, and individuals — that researchers and developers can tap into without paying cloud giants. - Public data commons: Curated, high-quality datasets spanning hundreds of languages and domains, designed to ensure AI systems reflect global diversity rather than just English-language internet content. - Cross-device compatibility: The organization says it has made "remarkable progress" running its AI stack across everything from high-end GPUs to low-power mobile devices.
The project has already demonstrated early prototypes running on consumer hardware, and Current AI says its next milestone is a public beta later this year that will let anyone contribute compute or use the network for free.
The bigger picture: Current AI joins a growing movement pushing back against AI centralization. Similar efforts include the open-source model ecosystem, federated learning initiatives, and government-backed AI infrastructure projects in Europe, Japan, and India.
Whether Current AI can scale its vision to compete with the trillion-dollar investments of Big Tech remains an open question. But its founders argue that the mere existence of a viable open alternative changes the incentive structure for everyone — and ensures that the future of AI isn't decided by a handful of boardrooms.
