This week has been a weird one for AI copyright lawsuits. Antrophic just won a suit to use copyrighted content, but pirating that content first is indeed illegal. Meta, similarly, won a case for its use of copyrighted content, but once again, there's a big asterisk.
As reported by TechCrunch, Meta won its case against 13 book authors whose work was scraped by its AI in a summary judgment. Effectively, this means the case was decided before going to a jury.
"This ruling does not stand for the proposition that Meta’s use of copyrighted materials to train its language models is lawful. It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one."
The central arguments made, according to Chhabria, are that "Llama is capable of reproducing small snippets of text from their books" and that Llama using the authors' work "has diminished the authors’ ability to license their works for the purpose of training large language models".
Chhabria calls both of these arguments "clear losers" as Llama doesn't generate "enough text from the plaintiffs’ books to matter, and the plaintiffs are not entitled to the market for licensing their works as AI training data". The one argument made by the plaintiffs that Chhabria does give credence to is that scraping their work could pump the market [[link]] with similar derivative work.
Despite suggesting this argument could have merit, the summary says the plaintiffs "barely give this issue lip service, and they present no evidence about how the current or expected outputs from Meta’s models would dilute the market for their own works."
Given that part of the plaintiff's argument is about them not being able to sell their books to LLM owners, I also don't buy it, as derivative work could still come from content willingly given to these tools.
There are four factors in determining fair use that judges consider. They are:
- The purpose and character of the use
- The nature of the copyrighted work
- The amount and substantiality of the portion used in relation to the copyright work as a whole
- The effect of the use upon the potential market for or value of the copyrighted work
Whilst clarifying that fair use is rather flexible and malleable to change, Chhabria focused on the fourth factor as the one that potential plaintiffs should attempt to win on.
"Because the issue of market dilution is so important in this context, had the plaintiffs presented any evidence that a jury could use to find in their favor on the issue, factor four would have needed to go to a jury."
Effectively, the claim made is that the plaintiffs not only misunderstood where they could win but also focused on arguments that weren't persuasive for the current application of copyright law.
"In cases involving uses like Meta’s, it seems like the [[link]] plaintiffs will often win". Chhabria continues, "It’s hard to imagine that it can be fair use to use copyrighted books to develop a tool to make billions or trillions of dollars while enabling the creation of a potentially endless stream of competing works that could significantly harm the market for those books."
Anthropic's recent case rules that it didn't break the law when using copyrighted material for its LLM Claude, but that using pirated material is still illegal. The judge in this case, Judge Alsup, argued Claude's use of that material is “quintessentially transformative."
Chhabria is critical of the notion that Judge Alsup focuses on the transformative nature of generative AI without acknowledging "the harm it can inflict on the market for the works it gets trained on".
In a similar vein, Getty has recently dropped a claim against Stability AI for its use of Getty's copyrighted material. Multiple cases are being settled each week around AI's use of copyrighted material, yet there appears to be room for a landmark case, according to Chhabria. In this specific case, though, "the plaintiffs presented no meaningful evidence on market dilution at all".
1. Best overall:
HP Omen 35L
2. Best budget:
Lenovo Legion Tower 5i
3. Best compact:
Velocity Micro Raptor ES40
4. Alienware:
Alienware Aurora
5. Best mini PC:
Minisforum AtomMan G7 PT