AI Narrated Audiobooks VS Copyright (April 26)
This month I’m talking about if you can copyright an AI narrated audiobook.
Video Image Description: There is an anime avatar representing Adara. Adara has purple hair and light blue bangs. She has white feathers on her head, a black collar with a heart emblem, and a smart blue and grey dress. She is standing behind a desk in an office. The ending screen shows a banner advertising Drawing Red, with a link to www.adaraspence.com/books. The paperback and ebook cover of Drawing Red shows a teenage girl drawing in a notebook with a large protective red and white wolf standing protectively behind her. It's set in a moonlit forest.
Transcript
Can you copyright an AI narrated audio book?
Well, yes and no. This is a deep topic, so grab a seat, put on your thinking hat, and see if maybe this is something you could pull off.
I'm Adara, a YA paranormal romance and sci-fi author, and I currently have an audiobook in production. I went down the AI audiobook narration rabbit hole for over a year, exploring one-click solutions like Google Playbooks as well as the current gold standard - quotation marks - ElevenLabs. But with video after video lording how good, cheap, and time-saving the process is, the question remains, can it be copyrighted?
In terms of the law, audio books are in a bit of a gray area, not unsimilar to AI generated book covers. When I say copyright for an AI narrated audiobook, I'm referring to that particular delivery of production of the story, not the text itself. For example, there can be different rights holders for audio to print editions. If you have a completely raw piece of text purely generated without any of your input, you have a different copyright problem on your hands, my friend. With that cleared, current laws state that for a piece of AI generation, there must be a significant input from a human. Quick side note, I'm based in the UK, but I'll be referencing US law because that's the home of Kindle Direct Publishing.
So, what does this look like? It does not involve using prompt after prompt for your generation. As the US copyright office has said, the reason you need so many prompts in the first place is because you don't know what creative decisions the AI is making and what the result will be with each click.
So, what does it mean? Editing, selecting, chopping up, rearranging clips, and post production work applied to the book are all pieces of human creativity that can be copyrighted. But the raw generated file itself cannot. Other examples include directing the tone, pacing, intonation, speed, and pitch of the delivery. Not all of the tools available, however, offer this level of input during or before generation in the process. This human creativity would need to be applied to a majority of the finished product. So, not necessarily the clean one button creation tool that it seems.
Whether you have permission to use a voice or even using your own voice clone, there's no distinction here. But of course, always make sure you have permission or the relevant licenses required. So if you receive an invitation from Apple, KDP or Google Books to create an audiobook at the push of a button, think twice. Especially for things like Apple Books, who don't offer you the ability to change anything about the generated book. You must alter the text itself, re-upload it, and have it regenerated again, which doesn't fit the involvement of creatively editing the audiobook.
It's a sticky area that I fear not a lot of people are taking into consideration in an environment where we're told audio is increasingly important, but the demand to produce and churn out material is high and budgets are low. Platforms and distributors have been quick to offer AI audio solutions to authors to get more books on their platforms to sell, but with little regard for the copyright of the author. Arguably, ElevenLabs offers you the ability to download your files and work on them externally, as well as having some level of ability to direct how it sounds before it is generated. But the threshold for editing and human input is high and their offer of a fast turnaround of a product will still leave a lot of people walking blind in the eyes of copyright law.
We saw this earlier in the author community where people started using AI to create book covers and the same principles apply now as they did then. How much of the cover was then changed after the fact by a human to allow it to be copyrighted? Ask yourself if your final product has undergone human creativity and how much. And in an age where there are increasing amounts of AI tools and solutions embedded into everyday software, what scope there is left for you to meet the threshold of human input.
Stay cautious and stay creative. Bye.
Credits
Live 2D Model: Akizone Customisable PRISM model
Ending theme and piano – Adara Spence and Mr Spence
Soon We'll Fly by Ghostrifter
Official https://bit.ly/ghostrifter-sc
Creative Commons — Attribution-ShareAlike 3.0 Unported — CC BY-SA 3.0
Free Download / Stream: https://bit.ly/35reep7
Music promoted by Audio Library https://youtu.be/q2vomZqSJuE
