BBC News
Getty imagesAn American judge has ruled that using books to train Artificial Intelligence (AI) software is not a violation of the US Copyright law.
The decision came out of a case brought by three authors last year against AI firm Anthropic, including the best-selling mystery author Andrea Bartz, who accused his cloud AI model to train and steal his work to manufacture a multiple dollar business.
In his judgment, Judge William Alsup stated that the use of anthropic of the books of authors was “highly transformative” and therefore allowed under the American law.
But he dismissed anthropic’s request to dismiss the case, the firm would have to test the use of pirated copies to build a library of his content to pronounce the verdict.
To prosecute with Ms. Bartz, whose novels we were never here and the last boat, the author of the Non-Fiction Writers Charles Grabber, The Good Nurse was: A True Story of Medicine, Madness and Murder and Kirk Walen Johnson who wrote The Fedar Chor.
An original company of Anthropic, Amazon and Google, a firm supported by Alphabet, can withstand up to $ 150,000 per copyright work.
According to the firm judge, it keeps more than seven million pirated books in the “Central Library”.
The ruling is one of the first people to weigh on a question that is the theme of many legal battles throughout the industry – how large language models (LLMs) can learn legally from existing materials.
“There is a desire to be a writer like any reader, the LLM of Anthropic has trained on works, not to run forward and repeat them or press them – but to turn on a hard corner and make something different,” Judge Alsup wrote.
“If this training process needs to make copies within LLM or otherwise, those copies were engaged in a transformative use,” he said.
He said that the authors did not claim that the training “violated the knockoff” with the replicas of their functions that arise for the users of the cloud tool.
If they had, he wrote, “This would be a different case”.
A similar legal battle has emerged from journalism articles to music and videos on the use of other media and content of the AI ​​industry.
This month, Disney and Universal filed a case against the AI ​​image generator Midzorney, alleging piracy.
BBC is also Considering legal action On unauthorized use of its content.
In response to legal battles, some AI companies have responded to the original material for license material for use, or striking deals with the creators of their publishers.
Judge Alsup allowed the “Fair Use” defense of anthropic, paving the way for future legal decisions.
However, he said that Anthropic had violated the rights of the writers by saving the pirated copies of its books as part of the “Central Library of all the books in the world”.
In a statement, Anthropic said that it was pleased with the recognition of the judge that the use of works was transformative, but some books were obtained and used.
The company said it was confident in its case, and was evaluating its options.
A lawyer for authors refused to comment.
