The research by Tomasz Hollanek and Katarzyna Nowaczyk-Basińska was published on 9 may 2024 in the journal Philosophy & Technology.

They explore the ethical implications of AI in the Digital Afterlife Industry (DAI), focusing on AI-generated simulations of the deceased, known as griefbots and deadbots or postmortem avatars.

Key Concept

AI technologies like griefbots offer new ways for individuals to process grief by enabling simulated interactions with deceased loved ones. These tools can provide comfort and a sense of ongoing connection, helping the bereaved navigate their emotional journey. However, the study emphasizes the importance of ethical considerations to ensure these interactions do not cause additional distress or emotional harm.

Key Findings

• Importance of explicit consent from data donors.
• Need for transparency in AI capabilities.
• Protective measures for vulnerable groups.

Key Elements

• Ethical concerns regarding postmortem privacy and human dignity.
• Impact on the mental health of service interactants.
• Recommendations for responsible AI development.
• Safeguards to prevent “unwanted hauntings” by AI chatbots.

Examples of Griefbots

• Project December: A platform allowing users to create AI-driven conversations with deceased loved ones.
• Dadbot: An AI created by James Vlahos to simulate conversations with his late father.
• BINA48: A humanoid robot that simulates a living person’s consciousness and memories.
• Eternos: The startup offers an AI-based legacy service, which allows you to communicate with loved ones even after you die. 
• Re;memory2: Korean firm DeepBrain AI launched the world’s first AI memorial service and communicate with loved ones after they died.
• Hereafter

For more details, visit Leverhulme Centre for the Future of Intelligence.

The full PDF of the research can be downloaded here.

Additional information can be found in this press release.

This research was published 9 may 2024 in the journal Philosophy & Technology.