Judge Rules That AI Training Doesn't Violate Copyright

[Argyle King]

There have been a few threads about AI on here. I felt that this would be of interest to the community:

Judge rules Anthropic did not violate authors' copyrights with AI book training.
(I originally quoted a different story, but the link wouldn't work. A different link is found at the end of this post.)

""The purpose and character of using copyrighted works to train LLMs to generate new text was quintessentially transformative,' Alsup wrote. 'Like any reader aspiring to be a writer."' ~Judge William Alsup

Judge rules Anthropic's training of AI with books is fair use https://share.google/hwNJzLkc3Nu9BCglB

Edit: Trying a different link

Federal judge rules copyrighted books are fair use for AI training

But, the judge ruled, AI companies shouldn’t be pirating the books they’re training on.

www.nbcnews.com

 

[Argyle King]

I'm having trouble getting the link to work.

The story has been covered by multiple outlets.

 

[Distracted DM]

Anthropic did not breach copyright when training AI on books without permission, court rules

Judge says firm made ‘fair use’ of literature but storage of pirated books in central library constituted infringement

www.theguardian.com

 

[Retros_x]

Great who controls now that AI companies buy all their books they feed into the AI as training data?

 

[MoonSong]

Great who controls now that AI companies buy all their books they feed into the AI as training data?

Nobody, they'll just have to buy them and save the receipts in case they get sued. And this is going forward because all existing current infringement is already done.

 

[The Firebird]

Seems reasonable in the case where they bought all the books rather than pirating them.

 

[MoonSong]

Seems reasonable in the case where they bought all the books rather than pirating them.

Which nobody has done n_n

 

[The Firebird]

Which nobody has done n_n

Part of the anthropic case was books that they purchased and scanned.

 

[MoonSong]

Part of the anthropic case was books that they purchased and scanned.

And many of these were bought after they had already pirated them which is one of the arguments given on the case still going on. Do you think Open AI and Google and Meta all bought copies to the works they used for training? (Contrast Adobe which has license to all material they used to train their AI tools)

 

[The Firebird]

And many of these were bought after they had already pirated them which is one of the arguments given on the case still going on. Do you think Open AI and Google and Meta all bought copies to the works they used for training? (Contrast Adobe which has license to all material they used to train their AI tools)

No.