Experts Alarmed by AI Generating Images of the Dead

Sandra Oparaocha

Writer

February 11, 2026
5:20 pm

A Danish Facebook group where members use AI to create images of deceased loved ones has sparked debate about ethics and mental health. Experts warn the practice could trap people in grief while raising concerns about consent and the spread of fabricated memories online.

Growing Trend of AI Generated Memories

A Facebook group dedicated to photo editing has become a gathering place for Danes seeking to recreate moments that never existed. The group, which now has over 63,000 members, allows users to request AI generated images showing them with deceased relatives or partners.

Personal Stories Drive Demand

Lene Brandt Sørensen lost her partner of seven years to a fatal blood clot just before they were set to marry. Nearly a year later, she turned to the Facebook group to fulfill a wish she never achieved while he was alive. She wanted a photo for her nightstand showing the two of them together.

The response overwhelmed her. Multiple members created images and videos using artificial intelligence tools, producing realistic depictions of the couple embracing. The results left her with mixed emotions, describing the experience as both wonderful and frightening.

Shift Toward Intimate Requests

Group members have noticed requests becoming more personal over time. While some people seek trivial edits like hairstyle changes or humorous additions to photos, others are asking for deeply emotional creations. These include images with deceased family members and recreations of moments stolen by death.

Benthe Johansen regularly helps fulfill such requests using paid subscriptions to AI platforms. She views her work as providing emotional support to people navigating loss. For her, creating these fabricated images offers comfort to those who might be grieving alone.

Psychological and Ethical Concerns

Mental health professionals and philosophers have raised alarms about the psychological impact of AI generated memories. The practice sits at the intersection of grief processing and technological capability, creating new questions about how people cope with loss.

Risk of Prolonged Grief

Bolette Windfeld Thesbjerg, a psychologist and PhD fellow at Aalborg University who researches AI use during vulnerable periods, warns of potential complications. Healthy grief involves two parallel processes: giving space to sadness while gradually accepting loss. AI generated images of the deceased could interfere with the second process.

The risk lies in becoming stuck in the grieving process rather than moving through it. When technology allows people to essentially bring loved ones back to life through images, it may prevent the acceptance necessary for healing. Though research remains limited, early observations suggest cause for concern.

Questions of Consent and Control

Philosophy professor Jens Christian Bjerring from Aarhus University emphasizes ethical considerations that grieving individuals should address. Before sharing images of deceased people in public forums, users should ask whether the person would have wanted this and how it affects other bereaved family members.

The open nature of the Facebook group compounds these concerns. When users share photos in a public space with thousands of members, they lose control over how those images spread. AI models continue training on data users provide, meaning elements of personal photos could appear in unexpected contexts later.

Broader Implications and Responses

The phenomenon extends beyond individual grief, raising questions about consent, data privacy, and the nature of memory itself. As the group continues growing by thousands of members weekly, the societal implications become more pressing.

Platform Growth and Scope

The photo editing group gained 2,400 new members in a single week in early February. Maja Nielsen, who lost her daughter in 2021, used the platform to create images of her grandson with his deceased mother. The emotional impact proved stronger than anticipated, highlighting both the appeal and the pain such images can evoke.

The response to these requests demonstrates widespread interest in the service. Nielsen’s post generated enormous engagement, with strangers offering empathy and technical assistance. However, the experience also prompted reflection about how widely she had shared intimate family images.

Need for Guidelines and Regulation

Experts agree that clearer rules are necessary as the technology develops rapidly. Currently, no regulations govern this market, leaving users vulnerable to unintended consequences. The AI models that create these images continue learning from submitted photos, but users have little understanding of where their data might surface later.

Thesbjerg stresses the importance of discussing consent as technology outpaces ethical frameworks. Without established norms, people risk causing more harm than good. Bjerring suggests the group establish clear ethical guidelines to prevent abuse and protect vulnerable users from crossing boundaries that could worsen their grief.

Mixed Results for Users

For Lene Brandt Sørensen, the AI generated images provided genuine comfort. Her post received 1,200 likes and over 200 comments, many offering condolences or edited photos. She now keeps two selected images by her bedside, finding solace when sadness strikes.

Yet the intensity of the emotional response catches many users off guard. Seeing a deceased loved one appear to come back to life creates powerful feelings that blend joy with renewed pain. The technology has advanced to a point where these fabricated moments feel nearly real, amplifying both their therapeutic potential and their psychological risks.