‘Heartbroken and powerless’: Wairarapa author caught up in Meta’s book copyright scandal

Friday, 28 March 2025

When fiction writer Pia Buck discovered several of her books had been stolen as part of an alleged multi-million book copyright scandal by tech giant Meta, she felt “exhausted and demotivated”.

The Carterton-based author’s titles were seven in more than 7.5 million books suspected of being copied by Meta, Open AI and other AI companies to feed and train their large language models (LLMs) — the algorithms that drive generative AI.

Court documents filed in the US described Meta’s actions as “massive copyright infringement” involving the download of a dataset of tens of millions of books and research papers from LibGen, one of the internet’s largest pirated libraries.

Lawyers for US organisation Authors Guild and affected American authors claimed Meta, which last year earned $165 billion, then fed the pirated works into Llama 3, the company’s latest LLM, “to extract … expressive content without having to pay”.

In a notice of motion filed with the Northern California District Court, it was claimed “Meta understood that the higher quality of the text, the better it could train its Llama models”.

Thousands of works by New Zealand authors were thought to have been infringed upon in the theft, which came to light last weekend when American magazine The Atlantic published an online search tool of LibGen’s database.

According to The Atlantic, Meta have argued that if its use of the LibGen dataset was challenged legally it could use a defence under the “fair use” concept. Under US copyright law, this allows somebody to justify the use of another’s copyrighted material for purposes including news reporting, teaching and research.

The independent author

For Buck, who works full-time as a writer, finding her titles, some of which were under her pen name Melissa Crosby, had been caught up in the alleged massive data-scrape was “a double blow”.

“Firstly, I don't have the resources to take Meta on. You've got authors in larger publishing houses that have filed lawsuits against them, but I can't do that.

“And on the emotional side of things, as an independent author I take on all the costs, all the losses, and it's hard enough for indie authors to get noticed in this landscape of authorship, and now we're having to fight against what could soon be an enormous library of AI generated books.”

While LibGen wasn’t the first online platform to pirate creative work this breach was especially tough to take, Buck said.

“To have the big guys like Meta cut you out like that — do you even bother fighting it? Honestly, it's exhausting and it is demotivating.”

While she acknowledged piracy was a part of everyday life, Buck was relying on her readers “to make the right decisions” and buy her work from legitimate sites.

The bookseller

The latest data-scrape of pirated material by big tech was another challenge in an already fraught industry, David Hedley, owner of Masterton’s independent bookshop Hedley’s said.

“It's hard enough making a living writing and book selling and publishing,” he said. “To think that all that effort could be undermined in such a way, I feel quite strongly about that.”

Like Buck, Hedley believed the wider community had a role to play in “standing up for creative people” to help counter the rise of AI content created on the backs of what the Authors Guild lawsuit describes as “professional fiction writers whose works spring from their own minds”.

“I think collectively, society needs to put pressure on governments and find any way possible to fight this, because otherwise it's catastrophic.”

Mary Biggs, operations manager of Featherston Booktown, agreed the rise of AI generated content — and the use of original human-made work to do it — presented something of an existential crisis.

“It’s great to have access to the wisdom contained in those millions of books,” she said. “But if it only enriches software companies and authors and publishers miss out, the books will stop coming and we'll end up a society of coders, not creators.”

‘It comes down to property rights’

Jenny Nagle, chief executive of The New Zealand Society of Authors (PEN NZ), was clear as to what Meta’s latest alleged actions constituted.

“This is theft. Meta is acting illegally,” she said. “Illegally digesting these works is copyright infringement under existing New Zealand law.”

Essentially, it comes down to property rights, she said, which was “something the Government understands”.

“Large corporations understand this. What they are doing [here] is deliberately taking and seeing what they can get away with. It is heartbreaking.”

Nagle said Meta could “afford to pay for their content”.

“What other manufacturer doesn’t pay for the ingredients of their products?”

Instead, writers were “paying the price”, Nagle said.

“Why are writers the ones paying for [Meta’s] access? Why are they the ones paying with their lost incomes for the training of these large language models?”

PEN NZ was compiling a database of New Zealand authors affected by the LibGen scandal and keeping a close eye on the Authors Guild suit in the US which if successful would “give us the legal backing for New Zealand authors and publishers to go directly to Meta and claim compensation”, Nagle said.

‘Dragging our feet’

As well as being alarmed at the latest data breach, Nagle was “enormously frustrated” with what she saw as the Government’s lacklustre response to the copyright threats posed by big tech.

The Ministry of Business, Innovation and Employment had not decided on the scope and timing of a review of the 1994 Copyright Act, which had stalled since it was started by the Labour Government in 2019.

Nagle said, “Why has [the review] not been fast-tracked to address this very important issue? Why are we dragging our feet?

“Are they happy for the property rights of New Zealanders to be taken away and usurped by some global giant?”

Changing NZ law would ‘not make positive difference’

Gillian Sharp, manager of corporate governance and intellectual property policy at MBIE, said the Ministry had been “closely monitoring overseas developments in copyright law” but said very few countries had implemented specific changes to their laws to address issues arising from the development and use of AI tools.

Because New Zealand copyright did not apply where New Zealander’s copyright may be being infringed by companies based overseas, “changing the New Zealand law would not make a positive difference”, Sharp said.

Within New Zealand, the Government was “neither allowing nor condoning the infringement of copyright”, she said.

‘Together we can all make change’

Sam Irvine, chief executive of the non-profit agency Copyright Licensing New Zealand, said the sensible thing for New Zealand to do would be to quickly follow and align with what its closest trading partners were doing to address the copyright challenges posed by AI.

“We don’t have the resources other countries have,” he said, “But generally the EU and Australia are working hard on this and as long as we're aligning with that, then I think we're in a good place.”

Ensuring responsible use of AI by public agencies and private individuals was also critical, Irvine said.

“Only use tools that will license content so that the creator is getting remuneration for that and has permission to use it.

“Basically, it's a protest, this is the way to make change,” he said. “There isn’t a lot in terms of regulation [here] that's going to have an impact on Meta or anyone else based in the US, but you can ensure that you're being responsible in the use of AI. Together, we can all make change.”