A disturbing trend has emerged from the conflict in Ukraine: Russian families are turning to artificial intelligence to create digital doppelgängers of fallen soldiers, raising profound ethical questions about grief, exploitation, and the limits of technology.
Startups in Russia are reportedly offering services that leverage machine learning and deepfake technology to generate interactive avatars of deceased loved ones. These avatars can mimic the voice, facial expressions, and even conversational patterns of the dead, drawing from photos, videos, and text messages. For grieving families, the appeal is clear: a chance to say a final goodbye, to hear a familiar voice, or to preserve a semblance of connection. But at what cost?
Critics argue that this crosses a fundamental moral boundary. ‘We are commodifying grief and normalising the use of AI to manipulate our most intimate emotions,’ says Dr. Elena Voskresenskaya, a Moscow-based psychologist specialising in trauma. ‘The technology is not ready for this. It can give false hope, distort memories, and prevent natural healing.’
From a technological perspective, the process is relatively simple: a large language model is fine-tuned on the deceased’s digital footprint, while a generative adversarial network (GAN) creates a synthetic likeness. The result is a customised chatbot that can hold conversations, tell jokes, or express love. Some versions even allow for text-to-speech synthesis in the deceased’s accent.
But the implications are staggering. Digital sovereignty is at stake: who owns the digital remains of the dead? In Russia, legal grey areas allow these startups to operate with little oversight. Families sign away rights to their loved ones’ data, often without understanding how it will be used. Once created, these avatars can be updated, shared, or even monetised without consent.
There is also the question of exploitation. Many families are desperate and vulnerable, willing to pay thousands of dollars for a chance to reconnect. Critics accuse these companies of preying on wartime grief, turning tragedy into profit. ‘It’s the ultimate Black Mirror scenario,’ warns Julian Vane, a Silicon Valley expat and technology ethicist. ‘We are building digital ghosts without understanding the psychological fallout. This is not a benign product; it’s an experiment on the human heart.’
Supporters argue that the technology offers solace and that individuals should have the right to choose how they mourn. Some cite religious or traditional practices that involve communing with the dead. But traditional rituals have boundaries; they do not confuse the living with the dead. An AI model can never truly replicate a person; it only creates a plausible imitation.
The broader context is also troubling. Russia’s war in Ukraine has already caused immense suffering, with tens of thousands dead. This technology threatens to extend that conflict into the digital realm, blurring the line between memory and reality. What happens when these avatars are used to spread propaganda or to manipulate families into supporting the war? The ethical perils are endless.
As the world watches, regulators scramble to catch up. The European Union’s AI Act touches on deepfakes but does not address posthumous avatars. In Silicon Valley, companies like Replika have faced backlash for allowing users to create romantic partners based on real people. But Russia’s war context adds a dangerous twist.
For now, the dead are being weaponised once more, this time by algorithms. The user experience of society has never been more fraught. We must ask ourselves: when AI can resurrect the dead, what does it mean to be human? And what responsibilities do we have to those left behind? The answers are not found in code, but in our conscience.
