The Rise of Pirate Audiobooks with AI Narration
On May 26, 2023, Gizmodo reported a surge in pirated audiobooks featuring AI-generated voices on YouTube. Author John Grisham has strongly criticized this trend, particularly YouTube's failure to remove such videos. Grisham stated that
“those who steal his works and profit from them should face both civil and criminal liability”.
YouTube responded to Grisham's concerns by explaining that the video remained accessible because no formal copyright complaint had been filed. YouTube spokesperson Jack Malon noted that the platform has spent over 20 years developing content protection systems, including Content ID, which is designed to detect pirated material. However, this system struggles significantly with AI-narrated audiobooks, as the technology alters the audio's sound, making it difficult to accurately match the track with the original.
Challenges for the Publishing Industry
Grisham also emphasized that
“YouTube is complicit in the problem because it knows about these videos but does not remove them”. Currently, a pirated video featuring an AI version of his novel 'The Widow' runs for 13 hours and has been listened to by over 80,000 people, according to The New York Times. The official audiobook of this novel was narrated by Michael Beck.
People choose AI-narrated books because they are free. Creators of such videos can make minor wording changes while keeping the content nearly identical, complicating detection and removal efforts. Publishers and authors are forced to manually submit takedown requests for pirated materials, adding further hurdles in the fight against copyright infringement.
The situation described by Grisham highlights the challenges facing the publishing industry in the digital age. The emergence of AI narrations threatens not only authors' incomes but also the integrity of copyright law. Despite existing content protection technologies, combating piracy is becoming increasingly difficult due to the rapid advancement of tools used to create pirated content.