Building the movement and infrastructure for AI in the public interest
The Collaborative for AI in the Public Interest is designed to spot and act on windows of opportunity for AI to deliver outcomes that matter to people and communities. Sometimes that means funding and championing organisations whose existing work is aligned with our mission, problem definitions, and strategies. Other times, it’s about identifying an unmet need or emerging opportunity that existing organisations are not well-suited to address and co-creating with entrepreneurs who have ideas ready to spring into action.
And then there are times when a problem needs to be addressed, there are no existing organisations working on it, and no one is ready with ideas for us to support. In those moments, we build. These circumstances drive us to ideation and require a functional approach – stepping in directly to fill gaps ourselves, for example by setting up an organisation or initiative. That’s what happened when Martin Tisné, CEO of The Collaborative, founded Current AI while serving as Thematic Envoy to the 2025 AI Action Summit in Paris.
Public interest AI: an idea in motion
In 2023, Martin Tisné was co-chairing the tech thinkers group convened by French President Emmanuel Macron when the early idea for a foundation for public interest AI began to take shape. By September that year, he had joined the Generative Artificial Intelligence Committee launched by French Prime Minister Elisabeth Borne.
The Committee’s March 2024 report to President Macron recommended establishing a public interest AI foundation. Martin was appointed Thematic Envoy for public interest AI, joining the Summit’s team, led by Special Envoy Anne Bouverot.
Over the following year, Martin conducted stakeholder consultations that shifted the idea from a foundation to a platform and incubator for public interest AI, where governments, philanthropy, and industry could collaborate to build the technical stack that public interest builders need.
That idea became Current AI, launched at the Summit in February 2025: a global public-private partnership with over $400 million commited in initial funding, including $100 million from The Collaborative, and the aim to mobilise $2.5 billion over five years.
Current AI shows what is possible when public and private focus and funding are mobilised behind a single goal: building open, public AI infrastructure at scale.
“We have a critical window to shape the future of artificial intelligence. We need to seize this window of opportunity before genAI business models are set in stone. Today, everything is to play for between a world where open models offer public interest options to all, and one where a few proprietary solutions prevail.”
Martin Tisné, CEO, The Collaborative; Founder and Chair of the Board, Current AI
One year on
The Collaborative has provided foundational support to Current AI since it was launched at the Summit, as a core funder on an ongoing basis and an operational steward during the organisation’s first year. Our team helped set up and run Current AI’s operations and programmes in its founding year, whilst we recruited a CEO.
Ayah Bdeir was appointed as CEO in January 2026 to lead Current AI’s secretariat, with Martin Tisné chairing the board.
Current AI’s governance structure
Current AI is headquartered in Paris with a global remit. It is structured as an independent non-profit organisation with a dual governance model:
- A Council, which convenes representatives from civil society, philanthropy, the private sector, and government to provide strategic input and exercise financial oversight.
- A Board, charged with fiduciary responsibility, including approving strategy, budget, and overseeing the Secretariat.
Current AI funds, builds and bridges
The AI systems that people interact with, such as chatbots, recommendation engines, and translation tools, are built on a stack: the underlying hardware, software, and data. A handful of private companies own and control that stack, giving them structural advantages: easier access to capital and resources and an ability to set a single, unified roadmap in pursuit of their own interests.
Current AI provides public interest builders that kind of advantage, not by attempting to compete dollar-for-dollar with big tech companies, but by funding, building, and bridging scattered efforts into a coherent and coordinated movement for public interest AI. The goal is a fully integrated stack that no single corporation or government owns or controls. Paired with funding, that stack makes it possible to develop and deploy public interest AI that private markets are not incentivised to invest in.
What this looks like in practice
Fund: In June 2026, Current AI established a pilot programme on cultural preservation, providing $3.2 million in funding to a grantee cohort that includes Masakhane (Kenya), Institute for Worldmaking (Lebanon), Portal sem Porteiras (Brasil), and African Internet Rights Alliance (Kenya), to support builders ‘who are creating open, high-quality, culturally grounded datasets, local models, tools and end-user applications to enable communities to make their own auditable, ethical and well-rounded AI systems’.
Build: The organisation then launched its AI Potluck methodology and roadmap, which calls for open source builders to bring and share pieces of the stack rather than building in silos. The AI Potluck’s first product is an ethical chatbot built in seven weeks entirely out of existing open source components.
Bridge: In July 2026, Current AI introduced its Open Source AI Gap Map, a living tool that visualises the state of the open source AI stack, the existing building blocks, their level of maturity, and the gaps that need to be filled.
Together, these efforts reflect Current AI’s nature as a builder-grantmaker that seeks to bridge silos and scale up the movement and infrastructure required for public interest AI to thrive.
“The public interest AI ecosystem lacks an independent, global space to nurture decentralised initiatives across the AI value chain. As a result, the benefits of scale and collaboration are often not accruing to the community doing this work. Current AI exists to change that by dedicating ourselves to building a truly open, accessible, and global AI system focused on the public interest.”
Ayah Bdeir, CEO, Current AI
Why it matters
AI should not be something that happens to us. It can be something that we – people and communities – build and govern together. Current AI is an early proof point that an alternative approach is possible, sparking imagination and new types of collaboration. Its first year of tangible results, from funding commitments to open-source tools now available to builders, shows that public interest AI is more than an aspiration.
Find out more
Read the problem definition behind our scale-up programming:
Read more
-
Problem definitionLack of infrastructure for public interest AI
The Collaborative 16 Jul 2026 -
Problem definitionOpportunities and threats within the AI supply chain
The Collaborative 16 Jul 2026 -
Problem definitionWeak public interest demand for AI
The Collaborative 16 Jul 2026