Openness has become an explicit subject across science policy and scholarly practice, where it is often vindicated in a rhetoric of optimism. In political discourse, as much as in the scholarly literature, open access to research data and publications is expected to enable what policy has typically failed to achieve by other means: that is, to overcome material, class, and political barriers that stand in the way of knowledge circulation. However, whether openness in science is a good thing or not also seems to depend on what is being opened, to what extent and for whom. In this paper I draw on different critical areas of Latin American science, technology and society studies (LASTS) to suggest that the current dominant views around open science can be limiting, as much as they could be enabling, more inclusive dynamics of access to and uses of scientific knowledge, especially in the peripheral (or non-hegemonic) contexts of science. These limiting views around openness, I argue, are linked with restrictive conceptions about science and its products: scientific activity is understood, by this token, as an invariably universal enterprise. In consequence, science outputs are conceived as self-contained knowledge products, and the processes and practices that account for their production and use are only partly taken into consideration. The aim is hence to elaborate on different forms of participation and exclusion to the processes of knowledge production which could help us understand how different stakeholders become engaged or excluded in the production of knowledge. To do so, I take the case of genomic research and drug development for neglected diseases as my empirical background. The argument draws on two concepts from LASTS. The first one is cognitive exploitation, according to which scientific outputs are used in for-profit contexts by third-parties, but without compensating the original producers. In this way, it is not only producers, users and appropriators of knowledge who become key in the dynamics of knowledge circulation, but also those acting as intermediaries. The other concept is integrated subordination, which refers, on the one hand, to the dynamics by which peripheral regions collaborate with elite research networks, and the difficulties that stand in the way of industrializing scientific knowledge, on the other. These difficulties spawn from the lack of capacities, but also from adherence to international research agendas, which are not necessarily connected with those required to attend to social needs in peripheral contexts. By putting into question the nature and the limits of openness, and by re-examining the types of knowledge at stake (beyond research data and publications), the actors, and their involvement, I suggest other ways in which open scientific knowledge could become effectively used.