AI ethicists have called for urgent safeguards against an emerging “digital afterlife” industry. The concerns centre on chatbots that mimic the appearances, speech, and personalities of dead people. Known as “deadbots” or “griefbots,” these AI clones are trained on data about the deceased. They then provide simulated interactions with virtual recreations of the departed. This “postmortem presence” can social and psychological harm, according to researchers from Cambridge University. Their new study highlights several risks. One involves the use of deadbots for advertising. By mimicking lost loved ones, deadbots could manipulate their vulnerable survivors into buying products. Another concern addresses therapeutic applications. The researchers...
