AI in Lucy Letby documentary: slop or preservation of authenticity?
Image by Cybernews.
The Investigation of Lucy Letby isn’t the first documentary to use deepfakes to protect vulnerable identities. Filmmakers say such digital avatars can help facilitate authenticity in ways that usual concealment techniques cannot.
The case of Lucy Letby, a British nurse convicted for life for murdering seven infants and planning to kill seven more, sent shockwaves across the country.
No less controversial were developments that followed the conviction, as a growing body of recognized experts now believe that the babies died of natural causes and there was no intentional harm.
The story is once again at the center of controversy after Netflix released a true crime documentary, The Investigation of Lucy Letby, on February 4th, 2024.
This time, however, discussions revolve around the use of artificial intelligence (AI) generated overlays to mask the identities of “Sarah,” one of the victim’s mothers, and Letby’s college friend, “Maisie.”
The decision to use AI to “digitally anonymize” subjects rattled both viewers and critics. A review from The Evening Standard described it as “a morally egregious use of AI slop,” while users on X called it “a terrible use of AI,” calling for it to be stopped immediately.
“Just hide their faces/distort their voices like a normal person. We don't need AI people,” another user wrote on X.
Is it common to use AI in documentaries?
While it may feel contradictory, AI is becoming more common in documentaries, despite their nature to portray real-life experiences.
Sometimes its use backfires, such as when the makers of the Anthony Bourdain documentary Roadrunner used AI to generate the chef’s voice to narrate his emails.
However, when the safety – or even life – of the movie’s subject is at stake, filmmakers see major benefits in using deepfakes.
Digital avatars were used in David France’s movie Welcome to Chechnya, following the lives of activists helping gays to escape the ultraconservative Russian republic, where they risk torture and killing either by family members or law enforcement.
France said it was crucial to ensure the highest level of identity obscuration due to the real risk of Chechen interviewees, who are LGBT persons, being hunted down even abroad.
Instead of entirely AI-generated faces, however, the movie “borrowed” faces of real activists in New York.
France described such “veils” in an interview as a way for his subjects to “narrate their story in the most intimate way while still protecting their anonymity.”
Facial expressions, posture, micro reactions, and eye contact are incredibly important – they’re fundamental to how you, the viewer, connect with contributors. In practice, when AI is used convincingly to replicate those physical cues, it helps to curate a stronger emotional bridge between the subject and the viewer.Cole Robinson
Cole Robinson, a documentary filmmaker at Filmit UK, says using AI has opened up the possibility of shooting interviews and scenes more “naturally,” without constantly working around concealment constraints, such as heavy shadowing and blurred faces, on set.
Most importantly, such AI use allows filmmakers to preserve emotional communication, because words are only part of how viewers interpret the human experience.
Robinson tells Cybernews, “Facial expressions, posture, micro reactions, and eye contact are incredibly important – they’re fundamental to how you, the viewer, connect with contributors. In practice, when AI is used convincingly to replicate those physical cues, it helps to curate a stronger emotional bridge between the subject and the viewer.”
Maintaining integrity is key
Another Body is one more notable example of what could be considered the ethical use of AI in documentaries. The movie follows Taylor, a college student, who is seeking justice after discovering that her pornographic deepfakes are circulating online.
Similar to Welcome to Chechnya, the movie uses AI to mask identities with “veils” made of real actors’ faces.
“I think that if we hadn't come up with this strategy of deepfaking her throughout the film to protect her identity, there is basically no way that she would have felt safe to do it,” the movie’s director and producer, Sophie Compton, said in an interview.
For some viewers, however, learning that a person they sympathized with in a movie was wearing an AI avatar may feel like “emotional betrayal.” They say they prefer silhouettes or blurred faces, which indicates that a real person is behind these obscuration techniques.
Robinson says AI can, in fact, facilitate the authenticity of the work if it enables vulnerable individuals to participate safely when they otherwise would not.
However, he emphasizes the importance of maintaining integrity, meaning that AI representations must faithfully mirror what was expressed and must be strict on not exaggerating or embellishing any emotions or reactions.
While he has concerns that the technology could become a tool for narrative manipulation in the future, he says that editing and post-production have always provided the ability to do so.
Robinson tells Cybernews, “The technology will continue to improve and become more seamless. Provided it is used with transparency and ethical intent, I think it represents a pretty valuable development in documentary storytelling.”
Unlock exclusive Cybernews content on YouTube.