Meta may have used books by Gerry Adams to train AI
Former Sinn Féin president Gerry Adams is among a number of authors whose books may have been accessed by technology company Meta to train their latest AI (artificial intelligence) model.
Mr Adams said the books had "been used without his permission", and the matter is now with his solicitor.
An investigation by The Atlantic magazine revealed Meta may have accessed millions of pirated books and research papers through LibGen - Library Genesis - to train its generative AI (Gen-AI) system, Llama.
A spokesperson for Meta said: "We respect third-party intellectual property rights and believe our use of information to train AI models is consistent with existing law."
The Atlantic magazine also published a database of books that had been pirated by LibGen, so many authors have been able to find out if their work appears on the site.
When BBC News NI searched the database a number of authors from Northern Ireland appeared on the list, including Jan Carson, Lynne Graham, Deric Henderson, and Booker prize winner Anna Burns.
Authors from around the world have been organising campaigns to encourage governments to intervene.
Meta, which owns Facebook, Instagram and WhatsApp, is currently defending a court case brought by multiple authors over the use of their work.
'Most lucrative act of theft in history'
Michael Taylor, a historian from Ballymena, said it is "infuriating" that Meta may have used his work.
Two of his books, The Interest and Impossible Monsters, both appear on the LibGen database.
"Writers spend years on their books, and contrary to what anybody thinks, very few people make enough money out of writing to live by their pen," he said.
"Meta might be worth more than a trillion dollars, and it might be politically untouchable, but by violating the copyright of so many thousands of books, its actions amount to the single greatest and the most lucrative act of theft in history."
Prof Monica McWilliams is an academic and former politician who has written extensively about the Northern Ireland peace process and domestic violence.
More than 20 of her academic papers and books appear on the database, including those on intimate partner violence and domestic violence against women during conflict.
She said when it came to her attention, she found it "quite shocking".
"The first principle in the academic world is that you direct your reader to your source material, and that isn't happening here," she said.
"It begs the question of what does copyright even mean anymore.
Prof McWilliams donates the royalties from sales of her writing to domestic violence charities like Women's Aid.
"If royalties are not being paid for the work to be used, then ultimately it is the charities that will lose out."
Last week, authors gathered in London to protest against Meta's actions and high profile authors including Kate Mosse, Richard Osman, and Val McDermid signed an open letter calling on the Culture Secretary Lisa Nandy to bring Meta management to parliament.
Posting on X, Richard Osman, who wrote the popular Thursday Murder Club series, said: "Copyright law is not complicated at all. If you want to use an author's work you need to ask for permission.
"If you use it without permission you're breaking the law. It's so simple.
"It'll be incredibly difficult for us, and for other affected industries, to take on Meta, but we'll have a good go!"
What is Llama?
Llama is a large language model, or LLM, similar to Open AI's ChatGPT and Google's Gemini.
The systems are fed huge amounts of data and trained to spot patterns within it. They use this data to create passages of text by predicting the next word in a sequence.
Despite the systems being labelled intelligent, critics argue LLMs do not "think", have no understanding of what they produce and can confidently present errors as fact.
Tech companies argue that they need more data to make the systems more reliable, but authors, artists, and other creatives say they should pay for the privilege.
887
