Meta has asked a judge in a landmark copyright case to throw out claims made by several high profile authors that it trained its AI mode, LLaMA using their copyrighted works.
The tech company stated: “Use of texts to train LLaMA to statistically model language and generate original expression is transformative by nature and quintessential fair use...”. It also pointed to Google’s use of books when it was creating its internet search tool.
It is appealing to the San Francisco federal judge presiding over the case, which opened in July, to throw out the claims that author Sarah Silverman, and others, have made – that Meta is abusing their copyrights.
The authors also sued ChatGPT - and therefore OpenAI - for infringement. AP News writes: “Silverman’s lawsuit says she never gave permission for OpenAI to ingest the digital version of her 2010 book to train its AI models, and it was likely stolen from a “shadow library” of pirated works. It says the memoir was copied “without consent, without credit, and without compensation.”” AI needs high quality language for optimum outputs; and therefore books are ideal.
Whilst only a handful of authors have launched legal wrangles, some huge names of the publishing world signed a letter accusing OpenAI, Google, Microsoft, Meta and other AI developers of exploitation of their works. Authors Nora Roberts, Jodi Piculit and Margaret Atwood were amongst the 4000 authors who added their signatures to a letter that stated: “Millions of copyrighted books, articles, essays and poetry provide the ‘food’ for AI systems, endless meals for which there has been no bill. You’re spending billions of dollars to develop AI technology. It is only fair that you compensate us for using our writings, without which AI would be banal and extremely limited.”
These copyright infringement case could prove of paramount importance as AI plays an increasing role in content production. OpenAI has not denied the use of Silverman’s work nor revealed its sources.
