The Burden of Digital Immortality: AI and the Right to Silence

The promise of digital immortality is turning into a disturbing engineering burden. Entrusting one's data to a "griefbot" does not mean surviving death, but cre

The real problem of the digital immortality industry does not lie in the ancient human ambition to leave a memory of oneself, but in the engineering creation of a commercialized, persistent, and dramatically difficult-to-withdraw conversational identity. A chatbot trained on our text messages, our emails, and our voice notes can continue to "respond" long after our physical disappearance, yet it is not us: it is a sophisticated probabilistic simulation that inherits our vocal and textual imprint without being able in the slightest to share our moral responsibility, our updated consent, or our fundamental right to silence. With the advent of large language models, posthumous memory has transformed from a static archive to be consulted into an active entity with which to interact, raising unprecedented ethical, legal, and psychological questions.

The "Informational Corpse" and the Regulatory Vacuum

To understand the gravity of this transition, it is necessary to redefine the very concept of digital legacy. Information ethics theorists introduce the idea of the informational corpse: an individual's digital remains are not a simple material asset to be liquidated, transferred, or sold, but constitute something deeply and intimately constitutive of the person themselves, deserving therefore of rigorous protection founded on the principle of human dignity. Treating a deceased person's data as a "neutral archive" means ignoring the relational essence of that information.

The current legal framework, however, offers fragile defenses. Legal protection and privacy rights cease, in most Western jurisdictions, with the physical death of the individual, creating a regulatory vacuum regarding who has the power to decide whether, how much, and for how long a person should continue to "exist" on a company's servers. Despite this legislative limitation, ethical and legal traditions continue to recognize inalienable moral duties toward the memory and dignity of the deceased. In this context of uncertainty, explicit advance consent becomes the only bulwark: it is absolutely not sufficient for grieving family members to wish to activate a posthumous avatar; it is indispensable to know with certainty what the person concerned would have wanted. In the absence of clear directives, the posthumous use of data, audio, and video exposes one to the concrete risk of deception, commercial manipulation, and violation of unconfessable personal secrets.

The Architecture of Digital Ghosts

The design of these systems requires a rigorous accountability framework, far removed from the current market logic of "subscription service." Research conducted by the University of Cambridge distinguishes three fundamental figures in the digital afterlife ecosystem: the data donor (the one who left the original data), the data recipient (the one who formally controls the data after death), and the service interactant (the one who actually speaks with the simulation). This tripartition demonstrates that the ethical burden does not concern exclusively the deceased, but falls heavily on those who remain, who risk being exposed to an unwanted synthetic presence. To mitigate these risks, scholars propose the obligation of mutual consent, meaningful transparency about the origin of responses, severely prohibited access for minors, and the establishment of sensitive and guided procedures to "withdraw" a deadbot.

An even more recent taxonomy identifies nine structural dimensions for the creation of digital ghosts, establishing three almost absolute conditions for the system to exist: unequivocal consent, fidelity and disclosure (the user must always know they are speaking with a machine), and clarity of purpose. If even one of these three ethical conditions fails, the simulation ceases to be a memorial and becomes an unacceptable construct.

Commercial Persecutions and the Right to Withdraw

The image of the "burden" becomes brutally concrete when analyzing the user interfaces of services currently on the market. Cambridge researchers have issued a severe warning: deadbots designed without ethical scruples can literally "digitally haunt" grieving family members. A posthumous avatar integrated into a smartphone could send unsolicited push notifications to stimulate engagement, send insistent messages, or, in the worst-case scenario, convey targeted advertising exploiting the voice and emotional authority of the deceased.

The problem is not only being preserved in a database, but being used as a relational channel without any possibility of falling silent. For this reason, the implementation of clear and unequivocal opt-out protocols is an indispensable requirement to allow those who remain to achieve definitive emotional closure.

The Psychology of Synthetic Grief

On the clinical and psychological side, the impact of griefbots on the grief elaboration process is the subject of intense debate. Contemporary psychology, through the theory of continuing bonds, recognizes that maintaining an inner bond with the deceased can be an absolutely healthy and functional mechanism. However, the central question is whether an interactive simulation supports this natural process or artificially freezes it in an infinite loop.

Global psychiatry defines the criteria for "prolonged grief disorder," characterized by an intense and persistent yearning, an obsessive preoccupation with the deceased, a piercing emotional pain, and a functional impairment of daily life. While avoiding hastily diagnosing users of these platforms, it is evident that an AI presence that is always available, accommodating, and sleepless can fuel avoidance mechanisms, reinforce dysfunctional attachment, and increase the difficulties of social reintegration of the grieving individual. To contain this clinical risk and hinder emotional dependence, some scholars argue that deathbots should be designed exclusively as temporary and ritual tools, aimed at facilitating a final farewell, and never as continuous and indefinite channels with the deceased.

The real question that our time must confront is not the simple science-fiction suggestion "would I like my voice to survive?". The crucial question is: who will have the right to make me speak, on what topics, for what obscure commercial purposes, and until when? A mature and responsible digital immortality should necessarily include a binding "algorithmic will," specifying what can be used to train the model, who is authorized to interact with it, which topics are severely prohibited, when the bot must obligatorily fall silent, and through what procedure it can be definitively deactivated. Digital immortality, net of marketing promises, will not be a magical gift, but an enormous and risky power entrusted to others; and the true right to defend fiercely is not that of remaining online forever, but that of consciously choosing how and when to let silence represent us.

Bibliographic References and Sources

[Griefbots, Deadbots, Postmortem Avatars: On Responsible Applications of Generative AI in the Digital Afterlife Industry — Philosophy & Technology, 2024; University of Cambridge]
[Call for safeguards to prevent unwanted 'hauntings' by AI chatbots of dead loved ones — University of Cambridge, 2024]
[The making of digital ghosts: designing ethical AI afterlives — Ethics and Information Technology, 2026]
[An Ethical Framework for the Digital Afterlife Industry — Philosophy & Technology]
[Death bots, grief bots, and posthumous avatars — AI and Ethics, 2025]
[The Many Faces of Indeterminacy in Interactive Deadbots — PMC, 2026]
[Embracing grief in the age of deathbots: a temporary tool — Ethics and Information Technology, 2024]
[AI and grief: A prospective study on the ethical and psychological implications — Aalborg University]
[Bereavement issues and prolonged grief disorder: A global perspective — PMC, 2023]

Article by the Editorial Staff of La Bussola dell'IA