Summary: Several British authors are calling on UK Secretary of State Lisa Nandy to hold Meta accountable for possible copyright infringement. Meta, allegedly, used LibGen, a database of 7.5 million books to train models, and the authors are heavily against that.
This is an argument that’s been going on for a couple of years, but it feels like an ages-old struggle. Content creators are at war with AI companies that steal their content for use in their AI models. Right now, several British authors want the UK to government to hold Meta accountable for allegedly using a database of stolen books.
Several creators are battling against AI companies for the same reason. Many of them came to the realizations that their content had been secretly scraped to train their AI models. Not only that, but these AI models can create content that can put these creators out of business. There’s the ongoing legal battle between The New York Times and OpenAI over this. We’re not sure when we’ll have some clear-cut laws regarding copyright and AI, but we know that the results from this case will set the precedent for cases going forward.
British authors want Meta to be held accountable for training its AI model on stolen books
A group of successful authors are now targeting Facebook owner Meta for possibly using pirated books to train its Llama 3 model. This is the company’s latest and most capable model at the moment. The pirated books live on a dataset called LibGen, and it consists of more than 7.5 million books. If Meta used this dataset, then it might be liable to sue for copyright infringement.
The Society of Authors, which includes authors such as Richard Osman, Kazuo Ishiguro, Val McDermid, and Sarah Waters, signed letter to the UK Secretary of State Lisa Nandy, urging her to go after the company. The goal is to push Meta senior executives to commit to respecting copyright law.
As put forth by the society, “These cases are shining a light on the unscrupulous behaviour exhibited by global tech companies which seemingly exploit copyright-protected material, safe in the knowledge that they will not be held to account.” This is not inaccurate.
At this point, there’s no telling what the outcome to this will be. Hopefully, if Meta did train its model on stolen data, it will face some sort of consequences for its actions.
