shakedown.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A community for live music fans with roots in the jam scene. Shakedown Social is run by a team of volunteers (led by @clifff and @sethadam1) and funded by donations.

Administered by:

Server stats:

255
active users

#FairUse

21 posts16 participants0 posts today

Anthropic did not breach copyright when training AI on books without permission, court rules
theguardian.com/technology/202
Judge William Alsup compared the Anthropic model’s use of books to a “reader aspiring to be a writer” who uses works “not to race ahead and replicate or supplant them” but to “turn a hard corner and create something different”.

Alsup added, however, that Anthropic’s copying and storage of more than 7m pirated books in a central library infringed the authors’ copyrights and was not fair use – although the company later bought “millions” of print books as well. The judge has ordered a trial in December to determine how much Anthropic owes for the infringement. #copyright #AI #fairuse #uslaw #uspol

The Guardian · Anthropic did not breach copyright when training AI on books without permission, court rulesBy Dan Milmo

It's not so much that #CreativeCommons is giving up in advance on the #AIcopyright issue—it looks like the decision-makers there sincerely believe that training is #fairUse, copyright should be limited, and so on

But most _users_ of CC licenses (authors, artists…) seem to have different beliefs—if an oligarch's murder robot is after me, I would pick up an "unclean" tool to fight them

IMHO we need an alternate license steward that's willing to better reflect user norms blog.zgp.org/fair-use-alignmen

blog.zgp.orgfair use alignment chart
More from Don Marti
Replied in thread

@mttaggart @AAKL I disagree.

True, "such as" denotes that the list of use cases is not exhaustive. But a proper #FairUse analysis should show that the use in question is one of those mentioned or a similar one ("such as").

The Anthropic judge does not take that step. He quotes section 107, but then continues as if that sentence was not part of section 107.

The Google Books ruling did better, pointing out the research and scholarly aspects of Google Books use.

Well...bugger. The link to the actual judgement is the first link in the article, but the article does a fair synopsis. Basically, the judge ruled against creatives and for the technocracy. I had hopes that Alsup would rule differently, since he understands tech better than most judges. I’m still parsing the ruling, but whereas I understand how he got to fair use, I don’t think it’s a correct ruling. I am not a lawyer, and I am especially not an IP lawyer, but this seems like it’s pretty much carte blanc to steal work from creatives.

techcrunch.com/2025/06/24/a-fe

TechCrunch · A federal judge sides with Anthropic in lawsuit over training AI on books without authors' permission | TechCrunchThe ruling isn't a guarantee for how similar cases will proceed, but it lays the foundations for a precedent that would side with tech companies over creatives.
Replied in thread

@mttaggart @AAKL Exactly, the judge contradicts himself.

First he quotes, as you point out, precedence (Google Books), saying that "the amount and substantiality of the portion used" does not 'matter so much', and then he says it actually matters quite a bit, because it favours (!) a finding of #FairUse .

He doesn't rule that it "doesn't matter much" (which would be a neutral result of the 3rd factor), he says it does matter, and in the opposite direction of what Fair Use has always been.

Replied in thread

@mttaggart @AAKL Now, the judge says that copying the entire work at least 4 times (additional copies, without destroying anything) means less copyright infringement, than if they had just copied a few sentences or pages.

Legally, that's absurd. It goes against all we know about #FairUse (which dates back to 1740, and in the current codified form to 1976).

This order screams for an appeal. 🙂

Replied in thread

@mttaggart @AAKL The ruling is very werid.

How can copying an entire book *favour* #FairUse?

One may come to opinion that it doesn't matter whether they scanned 1 page or the whole thing (not that I would agree). But to say that it would have been bad (for copyright reasons) if they had copied less is absurd.

Replied in thread

@Newk @heiseonlineenglish It is a really weird ruling. Inconsistent with precedence, inconsistent arguing in itself, and just obscure: How can copying an entire work *favour* #FairUse ?

Not that I support it, legally, but one could conceivable come to the conclusion that it doesn't hurt Fair Use to copy an entire work. But that taking everything at once favours Fair Use?

No way.

"A new paper from researchers at Stanford, Cornell, and West Virginia University seems to show that one version of Meta’s flagship AI model, Llama 3.1, has memorized almost the whole of the first Harry Potter book. This finding could have far-reaching copyright implications for the AI industry and impact authors and creatives who are already part of class-action lawsuits against Meta.

Researchers tested a bunch of different widely-available free large language models to see what percentage of 56 different books they could reproduce. The researchers fed the models hundreds of short text snippets from those books and measured how well it could recite the next lines. The titles were a random sampling of popular, lesser-known, and public domain works drawn from the now-defunct and controversial Books3 dataset that Meta used to train its models, as well as books by plaintiffs in the recent, and ongoing, Kadrey vs Meta class-action lawsuit.

According to Mark A. Lemley, one of the study authors, this finding might have some interesting implications. AI companies argue that their models are generative—as in, they make new stuff, rather than just being fancy search engines. On the other hand, authors and news outlets are suing on the basis that AI is just remixing existing material, including copyrighted content. “I think what we show in the paper is that neither of those characterizations is accurate,” says Lemley."

404media.co/meta-ai-model-memo

404 Media · Meta's AI Model 'Memorized' Huge Chunks of Books, Including 'Harry Potter' and '1984'Researchers found Meta’s popular Llama 3.1 70B has a capacity to recite passages from 'The Sorcerer's Stone' at a rate much higher than could happen by chance.