AI-powered chatbots are increasingly used in many contexts and for a variety of purposes. Among these uses, a particularly interesting one involves the so-called griefbots—that is, chatbots impersonating dead persons in the form of an artificial interlocutor. While they might help us process the loss of a beloved person, griefbots are not free from risks and may give rise to ethical concerns. This work aims at expanding the existing philosophical debate on griefbots. After providing a brief introduction to griefbots as well as the philosophical debate on their use, we provide an analysis of the way they may give rise to problematic ethical and normative tension between different transparency requirements. Then, we emphasise how the assessment of the related risks should adequately account for the dimension of users’ vulnerability.