This paper traces debates among standards-setters in China about accountability for the impacts of business activities and technologies (such as “AI”) on social and ecological systems. Standards are set by interlocking bodies—with different epistemological standpoints—who try to establish which data and which data-related technologies should be used to achieve accountability. Drawing on ethnographic examples from two standards-setting efforts by Chinese sustainable finance experts, this project finds that standards-setting reveals deep contestations about accountability, while simultaneously attempting to avoid them. This work extends existing sociotechnical research on standards by complementing studies of contestation among standards-setters with a focus on how systematic avoidance of this contestation among those same players shapes how standards are developed. Technocratic contestations are buried so deeply within technical systems that they are rarely visible to the outside. They are also rarely explicit, even to the actors involved. Yet, as this paper proposes, systematic attempts to avoid these contestations, while equally difficult to see, can exert substantial influence on how accountability (in this case, for environmental and social impacts) is measured, and therefore on the outcomes of accountability-related standards.