A bunch of Aussie authors have expressed outrage after discovering their books were used by Meta to train its AI platforms.
Making matters worse, it’s alleged that Meta — the parent company of Instagram and Facebook — gathered the books from Library Genesis (LibGen), an online database of books which pirates the titles rather than licensing them.
While the move is currently making waves Down Under, it actually kicked off in January in the United States, when American authors filed a lawsuit against Meta.
The lawsuit alleged that Meta owner Mark Zuckerberg approved the use of the notorious LibGen dataset to train the company’s AI models, despite being warned by his team that it is an archive “we know to be pirated”.
Aussie authors got wind of the controversy this week, when The Atlantic published a searchable database where writers can type in their name to see whether their work is included in the LibGen dataset.
A whole host of famous authors’ names appear in that database, none of whom were contacted by Meta to give permission for the use of their work for AI training (per ABC).
It’s also reported that no author who appears in the database was compensated by Meta.
Tim Winton, Holden Sheppard, Charlotte Wood, Trent Dalton and Helen Garner are among the Australian authors whose books were allegedly pirated by LibGen, as well as titles written by former prime ministers Malcolm Turnbull and John Howard.
In a statement, Australian Society of Authors CEO Lucy Hayward said writers “have quite understandably responded with great dismay and outrage at finding their life’s work has been pirated and used to train Meta’s AI”.
Hayward pointed out that Australian authors make just $18,200 each year from their work, making it “appalling that one of the world’s wealthiest companies has chosen to use creators’ work without permission or payment”.
She urged for government intervention and legislation to “safeguard creators’ rights” and “ensure AI developers are transparent, obtain consent for use of copyright material, and compensate creators fairly”.
For his part, Sheppard — whose hit YA novel, Invisible Boys, has been adapted into a series on Stan — was a little less diplomatic.
“[I am] fucking livid”, he told The Guardian.
“I am furious to learn my books have been again pirated and used without my consent to train a generative AI system which is not only unethical and illegal in its current form, but something I am vehemently opposed to.”
Meanwhile, author Sophie Cunningham told the ABC she is considering legal action, and Hannah Kent — whose entire three-book catalogue appears in LibGen — said she “feels a little like my body of work has been plundered”.
PEDESTRIAN.TV‘s own deputy editor, Matt Galea, discovered his 2023 book How To Spot The (Star) Signs was pirated by Meta, and had this to say:
“It’s a pretty grim discovery to find out your work has not just been stolen, but also used to support devious AI systems that will go on to steal other artists’ work.”
Earlier this month, Meta reportedly lobbied the Donald Trump administration to declare, via executive order, that training AI on copyrighted data, like that procured by LibGen, is fair use.
Lead image: @holdensheppard/Instagram and Chris Unger/Zuffa LLC
