"The Second Circuit held that Google’s book product did not engage in copyright infringement, neither for the literal copying of books to create its database nor for its “snippet” feature that allowed users to search a book for any passage and display that snippet in the context of the book without recreating the entire book. It is difficult to see how inference is not more transformative than this use."
This doesn't seem that difficult to me. A search engine is a means for finding new books, clearly different from the original. In contrast, generative models can be used to create works that compete directly with the works they were trained on—some image models even allow you to ask for works "in the style of" their training data. It's easy to imagine a court holding that this cuts against a finding of transformativeness, and even more important that this counts against the defendants for the "effect on the market" factor.
1. The issue with Google Books was yes creating copies for a search engine. But Google also *stored* those copies for a feature called "snippets" which would show you what you searched for in context. So they were displaying rather long segments of the book for free. That was also held to not be infringement.
2. You can't really copyright a style. The closest anyone got is the Marvin Gaye estate's lawsuit against Robin Thicke and Pharrell Williams, but that was effectively ignored in a subsequent lawsuit against Ed Sheeran where the Gaye estate lost because they couldn't copyright "basic building blocks" like chord sequences. This is why I think regurgitation remains the most potent potential issue.
3. I think the "effect on the market" factor will be the most important factor alongside transformative use, and it's interesting to ask whether a non-copyrightable thing (like a style) might be considered for an infringement factor (effect on the market). It will be interesting to see how courts weigh this.
"The Second Circuit held that Google’s book product did not engage in copyright infringement, neither for the literal copying of books to create its database nor for its “snippet” feature that allowed users to search a book for any passage and display that snippet in the context of the book without recreating the entire book. It is difficult to see how inference is not more transformative than this use."
This doesn't seem that difficult to me. A search engine is a means for finding new books, clearly different from the original. In contrast, generative models can be used to create works that compete directly with the works they were trained on—some image models even allow you to ask for works "in the style of" their training data. It's easy to imagine a court holding that this cuts against a finding of transformativeness, and even more important that this counts against the defendants for the "effect on the market" factor.
Some thoughts:
1. The issue with Google Books was yes creating copies for a search engine. But Google also *stored* those copies for a feature called "snippets" which would show you what you searched for in context. So they were displaying rather long segments of the book for free. That was also held to not be infringement.
2. You can't really copyright a style. The closest anyone got is the Marvin Gaye estate's lawsuit against Robin Thicke and Pharrell Williams, but that was effectively ignored in a subsequent lawsuit against Ed Sheeran where the Gaye estate lost because they couldn't copyright "basic building blocks" like chord sequences. This is why I think regurgitation remains the most potent potential issue.
3. I think the "effect on the market" factor will be the most important factor alongside transformative use, and it's interesting to ask whether a non-copyrightable thing (like a style) might be considered for an infringement factor (effect on the market). It will be interesting to see how courts weigh this.
4. Thanks for your comment!