Anthony Bourdain’s A.I.-generated voice has launched an important discussion in documentary filmmaking

Roadrunner, a documentary chronicling the life of the chef and television personality Anthony Bourdain, has garnered favorable reviews from critics who have described the film as “gripping,” “essential,” and “capturing everything about why viewers loved Bourdain.” It also sparked controversy over the use of artificial intelligence to generate a version of Bourdain’s voice that’s used to narrate a despairing email he sent to a friend. Is it okay to manipulate the voice of a deceased person who cannot give consent? Moreover, what are the ethics of doing so without alerting the audience that what they are hearing is actually a deepfake? The New Yorker‘s Helen Rosner, who first discovered the use of the A.I. tech in an interview with director Morgan Neville, reached out to two experts to address these questions. The first, Sam Gregory, the program director of a human-rights nonprofit that focuses on ethical applications of video and technology, suggested that the tech feels very new to us now but that we may become comfortable with its application as a filmmaking tool in a couple of years. The second, Karen Hao, an editor at MIT Technology Review, agreed that its use was new territory and that she would “forgive [Neville] for crossing a boundary that didn’t previously exist.” If anything, we can take comfort in the fact that though the audio was fake, Bourdain’s words were real. Sometimes the truth is messy.