What we’ve built so far

We began by creating a shared space and a core, interdisciplinary team. Before building anything, we defined non-negotiable boundaries through Clause 0, establishing clear limits on what this work would and would not engage in.

We then organized interdisciplinary sprints to explore concrete accountability pathways for AI, working from a shared backlog rather than predefined solutions. From this process, we developed two governance artifacts: an AI Harm Map prototype and a proposal for a systemic, escalation-oriented AI harm reporting mechanism, designed to surface patterns of harm and make governance failure visible.

The Alliance's work is carried by a distributed network of people spanning multiple regions and domains. Since launching in 2024, over 1,200 people have joined the initiative, 500+ have endorsed Clause 0, and dozens have actively contributed to governance sprints, applied research, and accountability frameworks. Contributors bring expertise from AI governance, technology, academia, design, human rights, law, and the arts. [Meet the people behind the Alliance →]

Existing AI governance mechanisms are largely reactive, fragmented, or inaccessible to those most affected by harm. This work exists to experiment with accountability where those mechanisms fall short.

Ethical AI Alliance did not begin with a product or a funding stream, but with boundaries, collective judgment, and the question of how harm can be documented and escalated in ways that institutions currently do not support. Funding enables continuity, coordination, and care for this work, rather than initiating it from scratch.