Digital Afterlife and AI Griefbots: Should You Create an AI Avatar of a Deceased Loved One?

Digital afterlife services use generative AI to create interactive chatbots, voice clones, or video avatars of a deceased person, trained on their texts, voicemails, emails, and social media history — an industry sometimes called "GriefTech," projected to reach a $5 billion market in 2026. Companies including HereAfter AI, StoryFile, and Eternos now offer these services commercially, and the technology has moved from a niche curiosity to a genuine consumer product category in the space of a few years.

This guide covers what these services actually do, the real psychological and ethical questions researchers and grief specialists have raised, the unresolved legal ownership questions, and a balanced framework for deciding whether this is something you or your family should consider.

What Digital Afterlife Services Actually Create

The output varies meaningfully by provider and price point. Some companies offer relatively simple text-based chatbots trained on a deceased person's written messages — texts, emails, social media posts — that can respond in a conversational style resembling how that person wrote. Others offer voice-cloned audio, built by analyzing recordings of the person's speech patterns, accent, tone, and pitch. The most advanced services offer interactive video avatars: StoryFile, for example, built a hyper-realistic video avatar of Holocaust educator Marina Smith that responded to questions with AI-enhanced, pre-recorded answers during her own memorial service in 2022.

Some services are built by a family after a death, feeding in whatever digital data the deceased left behind. Others are created proactively — a person "pre-records" themselves and their communication style specifically so a service can build a digital replica for their family to interact with after they're gone, a practice appealing particularly to people preparing to lose a loved one to a terminal illness or older parents wanting to leave something interactive for grandchildren they may not see grow up.

The Genuine Appeal, and Why the Market Is Growing

The appeal is not hard to understand. For someone experiencing sudden, unexpected loss — a death with no chance for a last conversation — the idea of being able to "talk" to a digital version of that person, even one built from old messages and recordings, offers something no physical keepsake can: a form of continued interaction rather than only static memory. This is a real and understandable emotional draw, and dismissing it outright misses why the industry has grown as quickly as it has.

The market reflects this: the digital afterlife services market is projected to reach approximately $5 billion in 2026, a figure that did not meaningfully exist as a distinct industry a decade ago. Coverage in Rolling Stone, The Conversation, and elsewhere in 2025 and 2026 has moved this from a fringe curiosity to a mainstream conversation about grief, technology, and what remembrance means in a digital age.

What Grief Researchers and Ethicists Are Actually Concerned About

The core concern raised by psychologists and ethicists studying this space is not that these tools are inherently harmful, but that they interact with grief in ways that are not yet well understood, and that the interaction may not always be healthy. Dr. Jessica Heesen, lead ethicist of the Edilife project at the University of Tübingen, has warned that continued interaction with a digital surrogate "could act like a painkiller in preventing the bereaved from accepting and dealing with their loss." Grief, in the traditional understanding most clinicians work from, has a natural arc that moves toward acceptance over time — and a digital surrogate that remains constantly, indefinitely available may interrupt that arc rather than support it.

There is also a documented "uncanny valley" effect — where an AI replica is close enough to the real person to be recognizable, but not quite accurate in personality, tone, or manner, producing a jarring, distressing experience rather than a comforting one. An AI memorial that gets a personality trait subtly wrong can function as a reminder of loss rather than a source of comfort, which is the opposite of what most people seek from these services.

The Emerging Security Risk: "Digital Kidnapping"

A newer and more concrete concern has emerged in 2026: reported cases of what has been described as "digital kidnapping," where hackers gain unauthorized access to a family's AI avatar of a deceased loved one and hold it for ransom, threatening to delete, corrupt, or misuse the digital replica unless payment is made. This is a genuinely new category of risk specific to this technology — an AI avatar built from years of a deceased person's private messages and voice recordings is a meaningfully sensitive dataset, and families using these services should treat account security for them with the same seriousness as any other sensitive financial or personal account, including strong, unique passwords and two-factor authentication where the provider offers it.

The Unresolved Legal Question: Who Actually Controls This?

As courts increasingly treat digital accounts as inheritable property, a distinct and largely unresolved legal question has emerged around AI avatars specifically: who has the legal authority to create one, and who can demand one be taken down or deleted? If multiple family members disagree — one wants to preserve and interact with an AI avatar of a parent, another finds the idea distressing and wants it removed — there is currently limited legal clarity on how that conflict is resolved in most states.

This connects directly to the pending federal NO FAKES Act, which would establish digital likeness as a formal, inheritable property right with a designated right holder — see our companion guide, The NO FAKES Act Explained. States including Washington, under its Forged Digital Likenesses Act, already treat a person's digital likeness as property that descends through a will, giving the estate clearer standing to authorize or prohibit this kind of AI use. See our Washington digital estate planning guide for details.

A Practical Framework for Deciding

If you are considering a digital afterlife service — either for yourself in advance, or for a deceased family member — a few questions are worth working through deliberately rather than deciding in the immediate aftermath of a loss, when judgment about long-term wellbeing is hardest to access clearly.

Would the deceased have wanted this? If the person is still living and this is being planned in advance, ask them directly and document their answer in their Letter to Family, the same place you'd document any other digital estate wish. If the person has already died without expressing a preference, consider what you know of their views on privacy, technology, and how they'd want to be remembered before deciding on their behalf.

Is this for early acute grief, or open-ended ongoing use? Many grief specialists suggest that time-limited engagement — using a tool during the hardest initial weeks, with an intentional plan to reduce or stop use as time passes — carries a different risk profile than unrestricted, indefinite interaction. If you choose to use one of these services, consider setting that boundary for yourself in advance, before grief makes boundary-setting harder.

Who else in the family needs a say? If multiple people share a relationship with the deceased, involve them in the decision rather than one family member unilaterally creating or maintaining an AI avatar others may find distressing to encounter. This is precisely the kind of family conflict our guide to starting difficult digital estate conversations is designed to help navigate.

What are the account security and data practices of the provider? Given the emerging digital kidnapping risk, ask any provider directly about their security practices, data retention policies, and what happens to the avatar if you stop paying or want it permanently deleted. Treat this decision with the same diligence you'd apply to any service handling sensitive financial data.

If You Decide Not to Use One — That's a Complete, Valid Choice

None of this is a recommendation to use these services. Many people process grief fully and healthily through traditional means — memories, photos, letters, conversations with others who knew the person — without any interactive digital component, and there is no evidence that this is a lesser or incomplete form of remembrance. The choice not to create or interact with an AI avatar of someone you've lost is a complete, sufficient decision on its own, not an absence of something you're missing out on.