In today’s algorithmic society, access to large-scale datasets is the sine qua non for any economic actor to reap the benefits of data-driven innovation (DDI). This article explores alternative mechanisms of data management in large-scale processing environments which can bolster access in view of the shortcomings of the existing data ownership-centric system. The scope of the analysis is limited to non-personal data. First, this contribution elaborates on the features and shortcomings of the data ownership-centric system and the existing legislation on data access. In fact, despite its ground-breaking potential, data access is not a widely available resource. It is subject, meanwhile, to the ability of several actors to control it, originating from data holders’ position of de facto control over data (“data ownership”), which is mostly anchored in technological, behavioural, and legal access barriers. This ownership-oriented setting thus stifles data sharing and opportunities for novel reuses of data. Despite these concerns, EU secondary legislation and case law (including the “essential facilities doctrine” of competition law) have not yet offered appropriate means to enable data access across society. Second, this article investigates whether alternative systems of data management based on the commons is a viable solution to open up access to raw non-personal data (RNPD). The commons as a conceptual notion and institutional mechanism values access and freedom to operate, instead of power to appropriate. The article homes in on two main reasons which substantiate why commons management of RNPD can be desirable. On the one hand, RNPD can be deemed a cooperative infrastructural resource that calls for being pulled out of its factual enclosure (“structuralist approach” of the commons). On the other hand, grasping RNPD as a commons means valuing its functional nature, making data available to a wide number of actors for the fulfilment of fundamental rights and enhancing human flourishing (“functionalist approach”). The article concludes with some thoughts on the lines of research which are still to be explored to put the commons-based vision of data management into practice.