Private-sector participation is already embedded in multilateral artificial intelligence (AI) governance. Yet it remains institutionally undefined and unstructured: who participates, why are they included, how is their input used, and what accountability follows?
This paper examines how participation can be structured so that firms provide the knowledge multilateral governance requires without allowing economic power to determine access, influence, or accountability. Drawing on survey responses from 32 private-sector actors, it explores uneven engagement, fragmented requirements, and the challenges of translating governance expectations into internal practice.
The findings also challenge a common assumption: a plurality of respondents consider that stronger mandatory accountability mechanisms would improve outcomes for their organizations and the public. While the sample cannot establish the position of the private sector as a whole, it suggests that assumptions of uniform business opposition warrant reconsideration.
The paper proposes three design principles: role-based participation, independently assessable reporting, and structural accountability. It outlines practical steps to move private-sector engagement from ad hoc consultation to an institutional relationship in which access, information, responsibility, and accountability are defined in advance.