Is it fair use when Google uses an entire scanned work in a machine learning project that leads to commercial gain?
That to me is a far trickier question...
That to me is a far trickier question...
https://en.wikipedia.org/wiki/Fair_use#U.S._fair_use_factors
Many "fair use" usages lead to commercial gain.
For example, "The third factor assesses the amount and substantiality of the copyrighted work that has been used. In general, the less that is used in relation to the whole, the more likely the use will be considered fair."
Using a work in its entirety surely must factor in.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2606731
TL;DR: "usage" by computer algorithm is quite different from showing a copyrighted work to a human.
There is a part re Sega v Accolade that brings up the interesting sub-topic of "intermediate copying", which is legal, but then says it's legal because it's “the only way to gain access to the ideas and functional elements embodied in a copyrighted computer program.”
But wholesale scanning of entire works is not the only way to gain access to the ideas of books. It's also possible to purchase copies from the publishers.
Perhaps a bit more relevant- "Accolade was not using Sega’s games for their protected expressive content, but simply to extract some unprotected, functional, nonexpressive information contained within them"
I think that would be debatable in certain machine learning projects. Perhaps much of it would fall under that category, but a large part of the point would in fact be to extract the expressive information contained in them.
What I'm trying to get at is related- Google (and other machine learning giants) is making an effort to extract the expressive content of these works. That's different from what we've seen in the past.
To me, at least, it feel like it's worth asking if it's right that the people who produced that expressive content should not be compensated for it.
Also, increasing discoverability through being searchable feels like it should pass under fair use. Can't make it searchable without a digital index, and an index can't be made without scanning!
(at least on paper)
>> Unlike other forms of Google search, Google does not display advertising to book searchers, nor does it receive payment if a searcher uses Google's link to buy a copy. Google's book scanning project started in 2004.
If that changes later on that may be cause enough to for them to be challenged on fair-use again.