As someone with astigmatism, I haven’t enjoyed the peak popularity of 3D movies and TV because I always felt dizzy and very uncomfortable trying to watch them. Since then, the idea of extended reality (XR) never really caught my attention.

However, during the lecture on XR and Immersive Analytics last week, I started thinking about an uncomfortable idea. Because I did not want to go off topic, I didn’t manage to raise it in class, but it has stayed with me ever since. This week’s article is about XR, AI, and digital life after death.

Apparently, this is not new — it is just me waking up to this idea with both awe and concern.

I had read before about people actively using Large Language Models (LLMs) and generative AI to chat with digital representations of deceased loved ones, but I wasn’t aware that even a couple of years ago this was already a business. People were already seeking help from AI‑generated avatars to process their grief after a family member passed away. At first, this was somewhat costly, but later AI cloning became affordable for most individuals.

According to an article from two years ago, “the market was particularly strong in China, where at least half a dozen companies were offering such technologies and thousands of people had already paid for them at the time.” The emerging technology is often referred to as “grief tech” or “griefbots,” and more broadly as the “digital afterlife industry” or “digital human industry.”

What I find interesting for the wider population around the world is that these tools are promoted as a way to process grief rather than as a need to continue the relationship. My concern here is how — or if — the “natural grieving process” will ever take place, the one we studied in books.

While we are more and more taught that negative emotions should be eliminated or pushed aside, I still believe that living through grief is the only way to acknowledge and respect what has occurred, and to learn how to live onward. For me, missing someone and crying over them feels more natural and true than pretending to have a conversation with them and continuing daily interactions as if nothing has changed — like in the Scarpetta episode.

There are also concerns about the vulnerability of users, the ease with which tech companies can commercialize grief (as they do with many other things), and the question of consent — especially if the deceased never agreed to have their personal data used in this way.

One of the articles I read described a U.S. court case last year in which an AI‑generated video of a victim forgiving his killer was presented during sentencing. The case was highlighted by experts as a troubling example of how digital recreations of the dead could emotionally influence legal decision‑making.

In a research paper published by Hollanek and Nowaczyk‑Basińska in 2024, recommendations targeted at providers of AI‑enabled re‑creation services “include suggestions for developing sensitive procedures for retiring deadbots, ensuring meaningful transparency, restricting access to such services to adult users only, and adhering to the principle of mutual consent of both data donors and service interactants.”

Since AI is built on data, if you don’t have good training data, your efforts may go to waste. For older generations, a lack of high‑quality data — especially videos and photos — may create limitations. But for most people today, we seem to be building our own datasets constantly, with millions of photos and gigabytes of personal data.

That brings me to the second question: would you ever intentionally want to build a digital copy of yourself to preserve your presence or continuity beyond death — in other words, to become digitally immortal?

From a technocentric point of view, creating a digital copy of yourself can feel like a logical extension of that worldview, offering a way to transcend biological limitations. But from my point of view, it is merely a copy — without consciousness. And if I sell my data and subscribe to a company’s service to stay “alive,” what happens if that company shuts down? How would I know that my data is controlled in a way I would consent to, and not exploited, misused, or sold?

At this point, I started wondering what we actually have in place in terms of policy and governance. Since I am not a scholar of social sciences, this is beyond my full capacity, but I still tried to look at the UNESCO and EU context to see whether governance is catching up with these developments.

UNESCO does not explicitly regulate digital immortality, but its Recommendation on the Ethics of Artificial Intelligence applies across the full AI lifecycle, mandating respect for human dignity, privacy, proportionality, and accountability. It explicitly warns against reducing people to extractable data beyond their lifetime and against the commercial exploitation of human identity, which can in practice constrain posthumous AI systems.

The EU AI Act does not prohibit posthumous digital avatars, but it imposes strict obligations when biometric data — such as voice, facial features, or behavior — are used. Many such systems are classified as high‑risk and subject to strong governance, human oversight, and transparency requirements.

Since GDPR largely applies to living persons, this framework reveals a recognized posthumous identity gap. At the same time, EU transparency rules require users to be clearly informed when they are interacting with AI rather than a real person.

Finally, just this month, we saw in the news that China is moving to tighten oversight of AI‑generated digital humans by introducing draft rules that mandate clear labeling, restrict emotionally immersive or addictive services for children, require explicit consent before using an individual’s likeness, voice, or personal data, and prohibit activities that infringe on rights such as reputation, privacy, and intellectual property.

What all of this reading and thinking leaves me with is not a clear answer, but a sense that we are moving faster than we are thinking — both personally and as a society. I can understand why someone would want to continue hearing their loved ones speak, or why someone might want to live on after death, but I don’t have a definite opinion about either of them at this stage, and I’m not sure if I ever will.

We are clearly entering a phase where death is no longer the end of data, identity, or interaction. Technology may be able to simulate presence, voice, and even personality, but it cannot — and perhaps shouldn’t — replace absence, loss, or consciousness. Whether that is comforting or disturbing may depend on where one stands.

Further reading:

• MIT Technology Review – Deepfakes of your dead loved ones are a booming Chinese business (May 2024)
• Forbes – Chinese Companies Use AI To Bring Back Deceased Loved Ones, Raising Ethics Questions (Juy 2024)
• Griefbots, Deadbots, Postmortem Avatars: on Responsible Applications of Generative AI in the Digital Afterlife Industry (May 2024)
• Nature – Ready or not, the digital afterlife is here (Sep.2025)
• The Conversation – An ‘AI afterlife’ is now a real option – but what becomes of your legal status? (Feb. 2026)
• Reuters – China moves to regulate digital humans, bans addictive services for children (Apr. 2026)
• China Daily – Draft rules to regulate digital human services (Apr. 2026)
• UNESCO – Recommendation on the Ethics of Artificial Intelligence
• The EU Artificial Intelligence Act

Initially published on Linkedin.