An alarming trend has emerged from the conflict in Ukraine: Russian families are turning to artificial intelligence to create digital simulations of loved ones killed in action. These ‘grief bots’, powered by generative AI, mimic the voice, mannerisms, and memories of the deceased, raising profound ethical questions that have now prompted a formal warning from the UK government.
London has labelled the practice a “digital ethics crisis”, with ministers calling for urgent international regulation. The technology, often built using open-source large language models and voice cloning tools, allows users to train an avatar on personal data such as text messages, photos, and audio recordings. The result is an interactive entity that can hold conversations, offer comfort, and even provide new ‘memories’ based on its training data.
The phenomenon is most acute in Russia, where the war has claimed tens of thousands of lives and state media has glorified sacrifice. For grieving families, the appeal is obvious: a chance to hear a lost son’s voice again or receive a digital hug. However, experts warn that these simulations risk preventing natural grieving, manipulating vulnerable users, and creating a dangerous precedent for digital immortality.
“This is the Black Mirror scenario we have been dreading,” said Dr. Elena Volkov, a digital ethics researcher at Oxford University. “These simulations are not the person; they are a statistical ghost. Yet they feel real enough to create emotional dependency. We are seeing grief commodified and exploited.”
The UK’s technology secretary has announced a review of AI-generated avatars, particularly those used for bereavement, calling for “guardrails to protect human dignity”. The concern is not limited to Russia. Similar apps have appeared in China and the United States, offering to ‘reconnect’ users with deceased family members.
From a user experience perspective, these tools are remarkably polished. They use natural language processing to recall specific events, adopt catchphrases, and even express affection. But the experience is fundamentally hollow. The AI has no consciousness, no continuity, and no ability to truly grieve. It is a mirror reflecting only the data we fed it.
The broader societal impact is disconcerting. As quantum computing advances, the fidelity of such simulations will only improve. We are approaching a point where the digital dead may become indistinguishable from the living in real-time conversation. This forces a critical question: do we have the right to digitally resurrect anyone, even with consent?
Russia’s stance on data privacy and AI regulation is notoriously lax, making it a fertile ground for such experiments. Some startups are already offering subscription-based ‘eternity’ services, storing a person’s digital footprint for future interaction. The UK warns that without international standards, these practices could normalise a form of digital slavery, where the deceased’s data is used for profit without ethical oversight.
For now, the crisis is contained. But as more families turn to AI for solace, the line between remembrance and replication blurs. The UK’s intervention is a critical step toward ensuring that technology serves humanity, not its grief. We must design for the living, not resurrect the dead.
