This presentation aims to reconceptualise the way in which we think about consent to the use of biomedical data research in the context of learning healthcare systems and population-level research more broadly. In the first half of the presentation, I illustrate the ways in which various kinds of ordinary permissions are multiply varied, involve different forms of engagement, different background assumptions and rely on different information. I suggest that the fundamental mistake in our current thinking about the ethics of biomedical data research (and elsewhere) is to assume that the model of permission-giving which applies in the clinical context is the model that ought to apply to all kinds of permission giving. This fundamental mistake I call ‘the fallacy of specific consent’. This observation and the avoidance of the fallacy allows us to view each type of permission as distinct and, specifically, as relating to governance in important different but justifiable ways. The second part of the presentation considers this sliding scale of permissions and governance and its direct relevance to the governance of research involving biomedical data. This allows us to see that permission-giving and governance should be thought of as working in tandem, combining together to help inform the processes by which individuals and society make decisions about participation in biomedical research. In the case of biomedical data research in the context of learning healthcare systems, this suggests a shift to governance and away from reliance on specific consent.