Imagine a farmer in a remote Indian village photographing a wilting crop, hoping to find out what’s wrong with it — except every AI tool she tries only understands English. That everyday gap is exactly what a nonprofit called Current AI is trying to close, one open-source language model at a time.
Founded in February 2025, Current AI has quietly become one of the most ambitious attempts yet to treat artificial intelligence as public infrastructure rather than corporate product. Backed by roughly $400 million in committed funding, the organization is building tools, datasets, and devices designed for the billions of people whose languages and cultures rarely make it into mainstream AI training data.
Key Highlights
- Current AI is a nonprofit founded in February 2025 to build open, publicly available AI infrastructure.
- The organization has secured roughly $400 million in funding from the French government, the Ford Foundation, the MacArthur Foundation, DeepMind, and Salesforce.
- Its CEO, Ayah Bdeir, previously led Mozilla’s AI strategy and founded the STEM company littleBits.
- Current AI co-built Suno Sutra, an offline AI device supporting 22 Indian languages, with India’s Bhashini language initiative.
- Last month, the nonprofit distributed $3.2 million in grants to four organizations in Kenya, Lebanon, and Brazil.
- It recently launched Alpha Chat, an open-source chatbot built in seven weeks with Hugging Face, Mozilla, and MIT Media Lab.
- Current AI is also partnering with Tokyo-based Sakana AI to build open AI tools for Japanese and Global South communities.
Quick Summary Box
What happened? A nonprofit called Current AI is building open-source, publicly available AI systems aimed at languages and communities that mainstream AI models largely ignore.
Why did it happen? Its founders argue that since every major AI system today is owned by a private company, a free public alternative is needed — much like the early World Wide Web.
Who is affected? Speakers of underrepresented and endangered languages, Indigenous communities, and developers who want to build on open AI infrastructure instead of proprietary models.
Why does it matter? Roughly half of the world’s spoken languages are at risk of disappearing, and English-centric AI development risks accelerating that loss.
What happens next? Current AI plans to expand its grant program, grow Alpha Chat, and deepen partnerships like the one with Sakana AI to build region-specific open AI stacks.
What Happened?
Current AI has been steadily rolling out projects since its founding, but recent months have brought a burst of visible progress. In February, at the India AI Summit, the nonprofit partnered with Bhashini — the Indian government’s AI language division — to launch Suno Sutra, a pocket-sized, offline AI device that works in 22 Indian languages without needing an internet connection. The device’s design is fully open-source, meaning outside developers can build on it freely.
Last month, Current AI allocated $3.2 million in grants across four organizations working on language and data projects in Kenya, Lebanon, and Brazil. And just last week, at the AI for Good Summit in Geneva, the group unveiled Alpha Chat, an open-source AI chatbot assembled in only seven weeks by a coalition of ten organizations.
Background
Current AI launched in February 2025 under the leadership of founder Martin Tisne, with a mission centered on public-interest AI. Ayah Bdeir, who joined as CEO in January after leading Mozilla’s AI strategy work, brings a track record in accessible technology: she previously founded littleBits, the STEM education company that reached millions of children before its 2019 acquisition by Sphero.
The organization operates as what Bdeir describes as a public-private partnership, pooling money from governments, companies, and philanthropic foundations to fund public-interest technology. The French government provided the initial $100 million in seed funding, joined by the Ford Foundation, the MacArthur Foundation, DeepMind, and Salesforce, bringing total committed funding to roughly $400 million. As Bdeir has framed it, the backers aren’t investors chasing returns — they’re funders supporting the mission itself.
Solving more than a language problem
Current AI’s leadership is careful to frame the project as more than a translation exercise. Bdeir has pointed out that language carries culture, memory, and tradition across generations — so when a technology can’t understand a community’s language, it also fails to preserve what that language holds. That framing shapes how the nonprofit selects and funds its projects.
Where the money is going
The first cohort of grants, worth $3.2 million, went to four distinct efforts:
- Masakhane (Kenya): building AI datasets covering more than 50 African languages, focused on health, agriculture, and education.
- Institute for Worldmaking (Lebanon): digitizing Arab cultural history into machine-readable databases that remain under community control rather than corporate ownership.
- Portal sem Porteiras (Brazil): developing offline AI tools with Indigenous Amazon communities, keeping the underlying data stored within their own territory.
- African Internet Rights Alliance (Kenya): creating audit tools designed to hold AI systems accountable across the African continent.
Who controls the data?
Data ownership sits at the center of Current AI’s approach. Bdeir has been blunt that control shouldn’t rest with a small number of companies concentrated in one part of the world. In practice, the nonprofit says its grantees store models and data locally, involve community experts before building anything, and build consent protocols into their pipelines so communities can pause or stop a project at any point.
Also read;
AWS Billing Bug Shows Some Customers Bills Worth Billions – Here’s What Really Happened
OpenAI Wants ChatGPT to Become Your Next Super App – Here’s What That Could Mean
Cognition Raises $1 Billion at $25 Billion Valuation as AI Coding Race Heats Up
None of the current grantees have fully solved the data-ownership question, and Bdeir doesn’t present that as a failure — she frames it as evidence that each project is treating the question seriously rather than letting a government or company decide unilaterally.
Building the open AI stack
Alpha Chat, launched at the AI for Good Summit in Geneva, represents Current AI’s most public technical milestone yet. Built in just seven weeks, the chatbot combined contributions from ten organizations, including Hugging Face, Mozilla, and MIT Media Lab, each supplying a different piece of the stack — from the underlying language model to safety tooling and computing power.
Current AI has also struck a partnership with Sakana AI, a Tokyo-based startup known for its work on what it calls Sovereign AI. Together, the two organizations plan to build a shared open-source AI stack supporting Japanese language and culture, while also extending to communities across the Global South that mainstream AI systems have largely overlooked.
Expert Analysis
Current AI’s strategy stands out for treating scale as a secondary concern. Rather than chasing the kind of massive user numbers that define Big Tech’s AI products, Bdeir has argued that success can look as modest as an Indigenous elder in the Brazilian Amazon using a Kenyan-built tool to pass ecological knowledge to the next generation in their own language. That’s a notably different measure of impact than the growth metrics driving most commercial AI development.
The nonprofit’s distinction between its own approach and Big Tech’s multilingual efforts is also worth noting. Bdeir has argued that large AI companies build multilingual capability mainly to expand their addressable market, often without meaningful consent from the communities whose language data gets used — pointing to instances where missionary Bible translations became training data for Indigenous languages before those communities had any say in the matter.
Why This Matters
Language loss is not a hypothetical risk. Roughly half of the world’s living languages face extinction in the coming decades, according to United Nations estimates. Because today’s dominant AI systems are trained overwhelmingly on English and a handful of other major languages, that imbalance risks compounding: languages that are underrepresented in AI training data become even less visible in the tools people increasingly rely on for information, education, and communication.
Current AI’s bet is that open, community-controlled AI infrastructure can slow that trend — not by replacing commercial AI systems, but by giving smaller and non-English-speaking communities the tools to build for themselves rather than waiting for a private company to decide they’re worth serving.
Impact on Communities, Developers, and the Industry
Language communities: Gain access to AI tools built specifically for their language and cultural context, often with direct input into how their data is used.
Developers: Get open-source building blocks — datasets, models, and devices like Suno Sutra — that they can adapt for local use cases without licensing restrictions.
The broader AI industry: Current AI’s public-private funding model and open-source releases could influence how governments and philanthropies approach AI investment going forward, positioning public-interest AI as a genuine alternative track alongside commercial development.
Timeline
| Date | Development |
|---|---|
| February 2025 | Current AI founded by Martin Tisne |
| January 2026 | Ayah Bdeir joins as CEO after leading Mozilla’s AI strategy |
| February 2026 | Current AI and Bhashini launch Suno Sutra at the India AI Summit |
| June 2026 | Current AI allocates $3.2 million in grants to four organizations in Kenya, Lebanon, and Brazil |
| July 2026 (AI for Good Summit, Geneva) | Current AI launches Alpha Chat, an open-source chatbot built with ten partner organizations |
| Ongoing | Partnership with Sakana AI to build a shared open-source stack for Japanese and Global South communities |
Important Facts Table
| Detail | Information |
|---|---|
| Founded | February 2025 |
| Founder | Martin Tisne |
| CEO | Ayah Bdeir (joined January 2026) |
| Total committed funding | ~$400 million |
| Lead funder | French government ($100 million seed) |
| Other funders | Ford Foundation, MacArthur Foundation, DeepMind, Salesforce |
| Flagship offline device | Suno Sutra (22 Indian languages, built with Bhashini) |
| First grant cohort | $3.2 million across four organizations |
| Open-source chatbot | Alpha Chat (built in 7 weeks with 10 organizations) |
| Key tech partners | Hugging Face, Mozilla, MIT Media Lab, Sakana AI |
Future Outlook
Current AI’s roadmap points toward expanding both its grant-making and its technical stack. The Sakana AI partnership signals an ambition to move beyond isolated regional projects toward a more connected, shared open-source infrastructure that multiple communities and countries can draw from. Given the pace of the past few months — an offline device, a multi-country grant round, and a chatbot built in seven weeks — the organization appears to be scaling its output even while keeping its funding relatively modest compared to commercial AI labs.
What Happens Next
Developers and organizations interested in Current AI’s tools can explore Alpha Chat directly, since it’s openly accessible, or look into the open-sourced Suno Sutra device for offline, multilingual AI use cases. Communities interested in applying for future grant rounds will likely need to watch for Current AI’s next funding cycle announcements, though the organization hasn’t detailed a fixed schedule for future grants.
Conclusion
Current AI isn’t trying to out-build OpenAI, Google, or Anthropic — it’s trying to offer something those companies structurally can’t: AI infrastructure that belongs to no one company and excludes no one language. Whether an offline device for Indian farmers, a dataset for African languages, or a chatbot stitched together by ten organizations in seven weeks, the common thread is the same bet Bdeir keeps returning to — that AI, like the early web, works best when it’s free, open, and built for everyone rather than owned by a few.
FAQ SECTION
1. What is Current AI?
Current AI is a nonprofit founded in February 2025 that builds open-source, publicly available AI infrastructure, with a focus on languages and communities that mainstream AI systems often overlook.
2. Who leads Current AI?
Ayah Bdeir serves as CEO. She joined in January 2026 after leading Mozilla’s AI strategy and previously founded the STEM education company littleBits.
3. How much funding does Current AI have?
Current AI has secured roughly $400 million in committed funding, led by a $100 million seed commitment from the French government, along with the Ford Foundation, MacArthur Foundation, DeepMind, and Salesforce.
4. What is Suno Sutra?
Suno Sutra is a pocket-sized, offline AI device built with India’s Bhashini language initiative that supports 22 Indian languages without needing an internet connection.
5. What is Alpha Chat?
Alpha Chat is an open-source AI chatbot that Current AI launched at the AI for Good Summit in Geneva, built in seven weeks by a coalition of ten organizations, including Hugging Face, Mozilla, and MIT Media Lab.
6. Who received Current AI’s first grants?
Four organizations shared $3.2 million: Masakhane (Kenya), the Institute for Worldmaking (Lebanon), Portal sem Porteiras (Brazil), and the African Internet Rights Alliance (Kenya).
7. How is Current AI different from Big Tech’s multilingual AI efforts?
Current AI’s leadership argues that large tech companies build multilingual models mainly to expand their markets, often without meaningful community consent, while Current AI focuses on community-controlled data and local ownership.
8. What is Current AI’s partnership with Sakana AI about?
Current AI and Tokyo-based Sakana AI are building a shared open-source AI stack supporting the Japanese language and culture, while also extending support to Global South communities.