The Unkillable Deathbot

AI “afterlives” (deathbots, generative ghosts) are typically imagined as platform services: hosted, moderated, and ultimately shut down by someone. This paper provokes a different future: a blockchained deathbot whose identity model, licensing, and revenue flows are enforced by smart contracts and decentralized storage. In this design, the bot can outlive companies, families, and jurisdictions, continuing to speak, earn, and possibly learn from conversations with no (minimal) human administration. We argue that “autonomy after death” is not only a technical affordance but a new socio-legal actor: part memorial, part micro-enterprise, part unkillable interface. Rather than proposing such systems, we use this scenario as a warning: a future in which an AI version of a person may continue to interact, transact, and resist removal even when families, platforms, or regulators would prefer it to stop. We outline emergent features, failure modes, and regulatory collisions (post-mortem privacy, consumer protection, AI governance, financial compliance), and we propose research questions and design provocations for DIS's “Beyond Interaction” theme.