Information systems scholars have been inferring data as ontologically unstable and epistemilogically uncertain and mobile. Data have been conceptualized as distinctive from digital technologies and possessing properties to relate with other data, digital technologies, actors and socio-political environments. Across such relations, data stabilize into larger objects, but also change as part of actors’ value-creation processes. However, data have been predominantly understood as open-ended, and the ability of data to simultaneously acquire structures and change has not been sufficiently explored – this requires an ontological investigation. The research question this paper seeks to address is “how can data, understood as both process and structure, be ontologically accounted for?”. The paper offers two contributions. First, it unpacks the process ontology of assemblage theory to account for data as dualities of structure and change. Second, it provides an understanding of data as irreversible historical productions which simultaneously engage in enduring and changing processes.