If civilization ever falls to a zombie apocalypse or nuclear Armageddon, we will need to have preserved centuries of accumulated practical knowledge to rise again. And if humanity should go extinct, leaving nothing but our legacy, the alien explorers who discover the ruins of our society would struggle to interpret human history without some great store of information to guide them.

Maybe these postapocalyptic scenarios are far-fetched, but even if society is never, say, drowned by the seas in some climate-driven disaster, leaving the remnants of humanity clinging to a few small bits of land, the massive collection of knowledge accumulated by the Internet Archive, comprising millions of books, is an invaluable resource.

By collecting and digitizing such a huge collection of works and lending them out online, the Internet Archive is making an incredible social contribution. The way the nonprofit manages that archive, however, has earned the wrath of book publishers. A few months into the coronavirus pandemic, when many physical libraries were closed, the IA began partnering with libraries to give users access to the IA’s collection, and removed digital limits on its lending. Several book publishers sued in June 2020, alleging a violation of copyright; the IA discontinued the practice a short time later.

Last month, a federal court sided with the publishers; Judge John G. Koeltl wrote that the IA had simply “copied the Works in Suit wholesale for no transformative purpose and created ebooks that … competed directly with the licensed ebooks of the Works in Suit.” The ruling went beyond this to say that controlled digital lending, or CDL, violates copyright law.

That’s significant, because for the past decade or so, many U.S. libraries have engaged in CDL, by which a limited number of digital copies, based on the number of physical copies a library possesses, are loaned out. Users lose access to these digital copies after a set period of time. The crux of the publishers’ complaint is that they want to charge libraries fees for ebooks, and they can’t do that if the Internet Archive is allowing those libraries to loan out its scanned copies for free. The ebook licenses that publishers sell to libraries, by contrast, have to be renewed after a fixed number of loans or a certain period of time, and they are highly profitable.

“The Publishers reasonably expect to be compensated for the reproduction of their copyrighted works,” Koeltl wrote, “and IA stands to profit from its non-transformative exploitation of the Works in Suit.” Although the IA is a nonprofit, Koeltl wrote that the IA benefits directly because it uses “its Website to attract new members, solicit donations, and bolster its standing in the library community,” and because it is paid whenever someone uses the website’s “Buy from Better World Books” button to purchase from the IA-affiliated Better World Books store.

The IA, in response, contends that it is doing nothing more than what libraries have traditionally done, loaning out copies of books it has bought. The only difference is that the copies are digital, rather than physical. But with CDL, the IA does not loan out more digital or physical copies than the IA has purchased. The Internet Archive also argues that there’s no evidence this lending has affected the publishers’ profits, which the judge concluded was irrelevant to the underlying legal matter. “Libraries have been around for thousands of years; they are older than copyright law itself,” the Electronic Frontier Foundation, which is representing the IA, wrote in its brief. “Never in the history of the United States have libraries needed to obtain special permission or to pay license fees to lend the books they already own.”