organizationNonprofit / AI Policy & GovernanceUnited States

Partnership on AI

Global nonprofit coalition advancing responsible AI for the benefit of people and society

Partnership on AI (PAI), formally the Partnership on Artificial Intelligence to Benefit People and Society, is an independent 501(c)(3) nonprofit coalition founded in September 2016. Its founding members included Amazon, Facebook, Google, DeepMind, Microsoft, and IBM, and it has since grown to more than 100 partner organizations drawn from industry, civil society, academia, and media — with more than half being human rights groups such as Amnesty International, the Future of Life Institute, and GLAAD.

PAI is explicitly not a trade group or lobbying organization. It develops tools, recommendations, and actionable guidance by convening diverse voices across the AI community, then works to drive adoption of best practices, inform public policy, and advance public understanding of AI's societal implications.

The organization's work is organized across five program areas: Safety-Critical AI; Fair, Transparent, and Accountable AI; AI, Labor, and the Economy; Inclusive Research and Design; and AI and Media Integrity. It is funded through membership dues from for-profit partners as well as grants and contributions from nonprofits and foundations.

As a multi-stakeholder body operating at the intersection of technology governance and civil society, PAI is highly relevant to any innovation touching responsible AI development, AI policy frameworks, or efforts to align emerging technologies with public interest outcomes.

Ideas that could help Partnership on AI

AI Model Ethical Opacity Disclosure Standard
A governance framework and accompanying software tool that helps AI developers document not what their model can explain about itself, but what it structurally cannot—turning Butler's philosophy of inevitable partial self-opacity into a practical, auditable AI transparency artifact.
Why it helps: This multi-stakeholder organization develops responsible AI norms; a Butler-inspired opacity standard would give them a concrete deliverable that bridges continental philosophy and practical AI governance.