Researchers, archivists, legal scholars, engineers, open-source leaders, and public-interest technologists working on: provenance · licensing · dataset governance · distributed training · archival preservation · open web infrastructure
Intention setting & project updates
Cross-group deep dives aligned with the three challenges
The AI Commons is a developing framework to support consent-aware, provenance-rich, open AI development.
It brings together legal tools, technical standards, and institutional partners to create a shared substrate for training, sharing, and governing models and datasets.
Rather than patching legacy copyright and data regimes, the Commons aims to define new norms around transparency, reuse, and reciprocal contribution.. norms that reflect how AI is actually built today.
AI development is accelerating faster than the social, cultural, and technical structures that support human knowledge. Existing copyright, data rights, and fair use traditions are downstream of a broader, longstanding societal conversation about how ideas are created, shared, and transformed.
We are now crossing a threshold in which the computational transformation of ideas demands upstream reflection.
This gathering starts from several simple premises:
These premises lead naturally to the need for transparency, accountability, and a commons of data, signals, and AI technologies accessible to all.
Our goal is to translate this into usable infrastructure:
This is an invitation-only working session (~20 people) focused on building early scaffolding for a durable AI Commons.
At this stage, the Commons is not a single license, dataset, or organization—it is a coalition of people working to architect the foundations.
This includes:
Together, we’re designing the shared infrastructure layer that modern AI has been missing.
To be discussed during our December Kickoff with intention of building something real & sustainable
How does it relate to existing copyright/data structures?
What would make it a de facto safe harbor?
Licenses, datasets, models, provenance specs, or all of these?
How do contributors benefit? How do we avoid enclosure?
Where is the line between transparency and control?
The goal is not to define “the Commons” as a philosophy, but to build it as an infrastructure that people can actually use.
Reach out if you're interested in getting involved (join our kickoff, community calls, meetups, etc)
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.
