In their interesting work, Goering and colleagues discuss the implications of next generation DBS in psychiatry on agency and identity and nicely illustrate the concept of relational agency. However, we identified three pressing aspects that fall short within the authors’ evaluations. First, we will argue that in psychiatry, pathological processes per se influence agency. We will exemplify this by the concept of loss of control in depression and include schizophrenia as the pathology of agency most research has focused on. Second, we will draw attention to the fact that in psychiatry, changing one’s personality is at the core of the therapeutic aim apart from highlighting the difficulty when wanting to investigate on what it means to “feel like yourself” in conditions characterized by chronicity and distorted self-images. With regard to the subjective experience of action, the article focuses entirely on potential effects of the neurodevice in reducing agency rather than including effects of hypo- and hyperagentic states, the neurodevice and their interrelation. As a corrective, we argue that in psychiatric hyperagentic states, neurodevices could even be aimed at reducing agency. Third, we remind the reader that for many psychiatric disorders symptoms are heterogeneous and lack a clearly identifiable neural correlate. This involves a serious signal detection problem that closed-loop devices for psychiatric disorders have to solve and that entails substantial risks for patients. We close by outlining the need to enrich metaphysical thoughts on agency by empirical research that e.g. include the intentional binding task.