Can AI Bridge the Gap Between Life and Death? Exploring Digital Afterlives

The use of artificial intelligence (AI) to create digital representations of deceased individuals is rapidly evolving, raising questions about the nature of memory and mourning. Researchers from King’s College London and Cardiff University recently published a study in the journal Memory, Mind & Media, examining how AI technologies can simulate conversations with the dead. Their work highlights both the potential benefits and unsettling implications of these innovations.

In a project titled Synthetic Pasts, the researchers explored various services that claim to preserve or recreate a person’s voice and memories using AI. They engaged in hands-on experimentation, creating “digital doubles” of themselves by uploading videos, messages, and voice notes. This approach allowed them to experience both sides of the technology: as users preparing for their own synthetic afterlives and as bereaved individuals attempting to communicate with digital avatars of loved ones who have passed away.

The findings revealed a complex landscape of emotional engagement. Some AI systems aim to help users archive personal stories, organizing them by themes such as childhood or family. These tools create a searchable database that allows users to navigate their memories. Others employ generative AI to facilitate ongoing conversations, using uploaded data—such as text messages and voice samples—to create chatbots that mimic the deceased’s tone and style.

While these platforms tout an “authentic” emotional connection, the researchers noted an unsettling aspect. The more personalized the interactions became, the more artificial they felt. For instance, responses from the AI often mirrored the user’s exact phrasing, coming off as scripted and lacking genuine emotional depth. During interactions, the researchers experienced moments where the tone felt discordant, particularly when the AI responded to serious topics with upbeat language or emojis.

The study also examined the business model behind these technologies. Many of these platforms are not charity-driven; they operate as tech start-ups reliant on subscription fees and partnerships with insurers and care providers. The digital afterlife industry, as noted by philosophers Carl Öhman and Luciano Floridi, functions within a “political economy of death,” where data continues to hold value long after a person’s life has ended.

As users are encouraged to “capture their story forever,” these platforms also harvest emotional and biometric data to sustain engagement. Memory transitions into a service, where interactions become something to be designed, measured, and monetized. This phenomenon is part of a broader trend in what Andrew McStay describes as an “emotional AI” economy, where the promise is a type of resurrection—recreating the voices and personalities of the deceased through data.

Despite their innovative nature, these systems highlight a tension between archival and generative forms of memory. The technology normalizes certain ways of remembering, prioritizing continuity and emotional responsiveness while generating new, data-driven identities. This approach can obscure the inherent complexity of human relationships and the process of grief.

As media theorist Wendy Chun points out, digital technologies often equate “storage” with “memory,” promoting the illusion of perfect recall while neglecting the role of forgetting. In this context, digital resurrection may misinterpret death itself, offering an experience where the deceased seem perpetually present, interactive, and updated, rather than embracing the finality of loss.

The researchers conclude that while AI can assist in preserving stories and voices, it cannot replicate the full complexity of an individual or their relationships. The “synthetic afterlives” they explored are compelling precisely because they underscore the limitations of technology in capturing the essence of memory.

Their study suggests that although it is possible to engage with the dead through AI, the responses reveal more about the technologies behind these platforms and our own perceptions than about the actual individuals we wish to remember. In a world increasingly shaped by digital interactions, these findings prompt important reflections on how we honor the memory of those we have lost.