The battlefields of Ukraine have produced a new kind of casualty: not of flesh and blood, but of memory and data. Reports from Russian media reveal a growing trend among bereaved families: the use of generative AI to create digital replicas of fallen soldiers. These “resurrected” avatars, powered by large language models and deepfake technology, can hold conversations, mimic speech patterns, and even offer comfort to grieving relatives. But as this practice gains traction, British technologists and ethicists are raising alarms about the morality of digital necromancy and the long-term psychological toll.
The technology is deceptively simple. Families upload photos, voice recordings, and text messages of the deceased. AI models then synthesise a virtual persona that can interact via text or voice, sometimes with a photorealistic avatar. One Russian start-up, reportedly called “Nekta,” offers a subscription service for £200 per year. The promise is a digital afterlife. The reality is a machine learning model trained on grief.
Silicon Valley expat Julian Vane, who now runs a tech ethics consultancy in London, warns: “We are downloading our loss into a black box. The algorithm doesn’t love you. It’s a statistical approximation of a human being. The danger is not just surveillance capitalism but emotional exploitation. These platforms collect intimate data, and soon enough, they’ll be nudging you to upgrade to the ‘premium’ version of your dead son.”
The ethical quagmire deepens when considering the nature of conflict. In Russia, where independent media is suppressed, the government has tacitly endorsed such tools as part of a broader narrative that glorifies sacrifice. But critics argue that offering a synthetic version of a soldier undermines the reality of death and may prevent families from moving through the stages of grief. The UK’s Data Ethics and Innovation Board has yet to issue guidance on “grief tech,” but pressure is mounting from MPs and mental health charities.
“We have seen the future, and it is a weeping mother talking to a bot,” says Dr. Aisha Khan, a psychologist specialising in digital trauma at King’s College London. “The initial comfort can be addictive. But there is a reckoning when the service stops, when the algorithm changes, or when the user realises the AI can never truly laugh or cry. That second loss is often harder than the first.”
The parallels to the Black Mirror episode “Be Right Back” are obvious, but the stakes here are geopolitical. Western tech companies have largely avoided this market, citing ethical concerns. But the vacuum is being filled by unregulated start-ups in regimes with looser oversight. This creates a digital sovereignty problem: whose values govern the simulation of a human soul?
Vane suggests that Britain must lead on this issue, drawing a line between therapeutic tools and exploitation. “We cannot legislate for grief, but we can insist on transparency. Every AI-generated interaction should carry a label: ‘This is not real. This is a machine.’ And we must consider a ‘right to digital death’ — the ability to delete a persona when the user is ready.”
The story is not just about Russia. It’s a mirror for our own culture. As AI becomes more human-like, the boundary between memory and simulation blurs. The technology is moving faster than our ethics, and the casualty may be our collective mental health. The question is not whether we can resurrect the dead, but whether we should. And who will be left to answer when the algorithms do the mourning for us.
