So, I got an email from the woman who runs my writing group complaining that the Internet Archive had illegally digitized "thousands" of books (I suspect that that's off by an order of magnitude), and was now making them available for free under the guise of fair use. She also scoffed at the idea that the Archive will compensate authors, as it claims it will.
Naturally, I had to look this up. This WIRED article and this Smithsonian article give a pretty good overview of the controversy. I was going to link the library itself, but realized that if Internet Archive is engaging in copyright infringement (and it likely is), that could constitute contributory infringement on my part, because US copyright law is whack. I will, however, link to the FAQ, which provides the Internet Archive's take on this.
I will also take this opportunity to link y'all to Tom Scott's excellent mini-documentary: No copyright infringement intended. A link because I don't think y'all want to watch forty minutes of embedded video in one go. Scott goes over the way copyright works in the US legal system, and makes pretty clear that, from a legal perspective, the Internet Archive is 100% in the wrong here.
Legally speaking, it's not even close. The Internet Archive may have the right to digitize these books—that part isn't clear—but they do not have the right to make copies of those digital scans and share them with others without the permission of the copyright holder or someone (such as a publisher) who has sublicensing rights. US courts have sided with copyright holders in far less ambiguous cases than this one.
And no, this isn't fair use either. Like, if I were on a jury, I would bend over backwards to find fair use while plausible, but this isn't remotely fair use. The most important factor for fair use is the whether a work is transformative, which this clearly isn't. The other three factors are the amount used (Archive uses the whole works), market usurption (Archive focuses on out-of-print books but doesn't limit itself to those), and whether it's commercial or not (the only factor on the Archive's side here).
From a pragmatic perspective, the long-term best-case-scenario here is that the Internet Archive ends up paying out vast sums of money, and getting a license to use a small handful of the books it's digitized—mostly the ones that are already available through other sources anyways—going forward. But the worst-case scenario is that in five years' time, the Internet Archive won't exist because of this, and I'd say that this scenario is far more likely than the best-case scenario. Indeed, technically, even the Wayback Machine itself is a massive experiment in copyright infringement under current US law, which to my mind is more a problem with US copyright law than the Internet Archive, but it's not anything I can do about.
So this raises the final question: for whom is the Internet Archive's National Emergency Library desirable? Obviously, anyone who borrows books from it that there would have been a waitlist for under it's old (also blatantly-illegal) will benefit. Publishers and authors could stand to benefit in the long-term, and I'm skeptical that this is hurting them much in the short-term. The Internet Archive's ebooks require you to create an account and are subject to DRM, which means this is not a service I would ever use. I feel like advocates of DRM should therefore be over-the-moon that even the Internet Archive, advocates of openness, are using their technology. From my perspective, as someone who is strongly opposed to DRM, this is conversely a bad thing.
In short, I think that this is a decision which probably won't help nearly as many people as the Internet Archive thinks it will, won't hurt authors and publishers nearly as much as author's organizations and publishers fear, and will probably end up being a major self-inflicted wound in the cause of openness, since the most likely effects of this will be A. to expose more people to the idea that ebooks should be DRMed by default and B. to harm or even shutter the Internet Archive.