The use of artificial intelligence (AI) to simulate the presence of deceased individuals in the form of virtual conversational agents—often referred to as “deadbots”—raises complex ethical questions. Deadbots, holograms, conversational agents… Is AI stronger than death? This is one of the main questions addressed during the 14th European Forum on Bioethics from February 7-10 in Strasbourg, France.
The growing possibility of virtually reviving deceased individuals to alleviate the suffering of grieving people raises numerous questions regarding the respect for privacy and consent of the deceased person, the emotional impact created on the remaining living beings, the concept of death itself, or the indispensable questions of AI regulation.
Privacy and Consent
According to some of the speakers at the 14th European Forum on Bioethics, it is crucial to determine whether the deceased person would have consented to this use of their data while alive, and how to respect their privacy even after their death.
“During their lifetime, a patient may request to be filmed or have their voice and other memorabilia captured through digital recording tools. However, this raises questions of image property rights and consent. Who will have the right to use these images or reconstructed words once the person has passed away?
Furthermore, this tool would need to be composed of memorabilia from the person, but who will decide on these memorabilia if not the deceased person themselves? What if it’s someone who dislikes this person who wants to reconstruct something? We end up with a perversion of the deceased person’s image, a denial of their self, so if we want such a tool, it must be built with an ethics that absolutely considers the wishes of the deceased individual.
Moreover, in some cases, not all members of a family have the same attachment to deceased individuals. Some members may not have seen these individuals for years. There may be disputes over burial, cremation, body donation to science, etc. For the question of ‘deadbots’, we may face the same problem. So in the will, should we not anticipate the issue of our post-mortem image? There is already a discussion today about virtual cemeteries or virtual mortuary sites.
I believe that tomorrow we will have to invent a new law in funeral legislation.”
Emotional and Psychological Impacts on the Grieving People
The use of deadbots can have a significant emotional impact on grieving individuals. It may offer some comfort to some, but others may perceive it as disruptive or even psychologically disturbing. According to Fiorenza Gamba, Socio-anthropologist at the IRS, University of Geneva:
“Deadbots raise difficulties and can in some cases plunge the still-living into an inability to move on from mourning. Others may consider this tool as a means to bid farewell to their loved ones who passed away too abruptly or at a young age.
Mourning is a very personal experience that follows individual rhythms, so the important thing is to ensure that vulnerable individuals are not exposed to the risk of being trapped by this technology. This is an important aspect that all efforts to establish AI regulation must take into account.”
A New Conception of Death and Afterlife?
On the other hand, the use of deadbots can also influence our conception of death and afterlife. It may alter our traditional understanding of grief and the memory of the deceased. Grégoire Moutel, who manages the mortuary chamber of the University Hospital Center (CHU) of Caen in France, highlighted:
“In the collective imagination, forensic medicine is associated with the medicine of death. In forensic medicine, we often say that we ‘make the dead speak’ because we reconstruct a story after death that was not witnessed during the person’s lifetime.
As early as the 18th century, people practicing spiritualism were already making the dead speak. It wasn’t just a game; they believed in it, and it worked. Questions were already being asked at that time: Are these people twisted? Are they mystifying death? And does it benefit the living?
Today, we must ask ourselves: in medicine, there is what we call ‘pathological mourning.’ Some people will go to the cemetery every day and feel fine, while others will go to the cemetery every day and not feel well. Ultimately, it’s not the tool that will make us feel good or bad, but the nature of the individual. Therefore, we must accompany the use of these deadbots so that they can be tailored to each person, rather than banning them because, anyway, it’s pointless to impose bans as there will always be ways to circumvent them.”
The Death of Death?
“Do these tools mark the death of death? Computer or virtual immortality can be imagined in three ways:
1. The merging of man and machine (which is Elon Musk’s dream, to create something that is both a computer and a flesh being with connections, without aging or suffering).
2. The possibility of reproducing our consciousness and putting it on a computer medium, so that we can continue our life on a computer; some specialists believe this is entirely feasible.
3. The third hypothesis would be to have virtual beings that are the continuation of physical beings but reproduced on a computer medium, essentially having virtual beings that are living copies of deceased individuals.
All sufficiently advanced AI is indistinguishable from consciousness. Some extremely competent and intelligent people, especially at Google, already believe that AI is conscious… If AI allows us to live much longer, in good health and without disease, isn’t this the best we can hope for from this technology?”
Responsability and Regulation
According to the speakers, in this context it is crucial to establish clear regulations and guidelines to govern the use of deadbots, ensuring the protection of individual rights, transparency in their development and use, as well as accountability in cases of harm or abuse.
Maria Fartunova Michel, holder of the Jean Monnet Chair in EUBioethics and Associate Professor with HDR in public law at the University of Lorraine / IRENEE, said:
“There is legal uncertainty surrounding the regulation of AI that is still under discussion. Digital reality and AI are now part of the reality of all domains, and more than ever, the demand for law is necessary.”
According to her, in the face of AI’s promise to defy death, it is important to recall and maintain “the freedom to dispose of one’s body after death.”
But what about the legal status of this virtual person? Maria Fartunova Michel believes “that AI should retain its status as a ‘thing’ and not be tied to the status of a human being.” According to her, this entire debate questions the individual’s relationship to death and is an original way of rethinking bioethics as a whole.