Mar 18, 2026

Tomasz Hollanek, a researcher at the University of Cambridge, says AI bots that are coded to mimic a deceased person could help preserve historical figures. But, if developed unethically, a late person's data could be manipulated to benefit a business model.

Jake Lomachevsky/Getty Images

There's an emerging industry that uses artificial intelligence to create simulations of people who've died. These post mortem avatars are also called griefbots.

Some critics, including Tomasz Hollanek, a researcher at the University of Cambridge, say this practice raises a number of ethical issues. He walks us through the mechanics of how this technology works, and how it may or may not be used responsibly.

More on this

“Family deepfakes help people celebrate and grieve in India” from Rest of World

“Griefbots, Deadbots, Postmortem Avatars: on Responsible Applications of Generative AI in the Digital Afterlife Industry” from Springer Nature