Books Legal Issues Library Issues Technology

Adobe Digital Editions and infoleaks

Eliminate DRMThe online library world exploded today over the revelation that Adobe Digital Editions, software that is required for many library-focused eBook services, evidently leaks like a sieve when it comes to our user’s information. The TL:DR version of the story is that ADE appears to be sending in plain text to Adobe’s servers information such as: the book you are reading, title, publisher, which pages you have read and which page you are currently on. Much longer discussions about the leak and potential fallout here:

Andromeda and Galen then both went on to touch on some of the core problems with this leak, focusing on the conflict between Adobe’s action and the ethics of librarianship, and the possible role that ALA may have in bridging the gaps in libraries’ knowledge of these actions.

There are a few things I wanted to emphasize about this situation. The first is that several of the reports have noted that earlier versions of Adobe Digital Editions didn’t seem to “spy on its users” in the way that the most recent version (version 4) does, and recommend using earlier versions. The truth of the matter is that of course the earlier versions are spying on users…they just aren’t doing it in as transparent a manner as the current version. We need to decide whether we are angry at Adobe for failing technically (for not encrypting the information or otherwise anonymizing the data) or for failing ethically (for the collection of data about what someone is reading). It’s possible to be angry at both, but here’s a horrible truth: If they had gotten the former right and encrypted the information appropriately, we’d have no idea about the latter at all.

I think that Andromeda has it right, that we need to insist that the providers of our digital information act in a way that upholds the ethical beliefs of our profession. It is possible, technically, to provide these services (digital downloads to multiple devices with reading position syncing) without sacrificing the privacy of the reader. For example (and this is just off the top of my head) you could architect the sync engine to key off of a locally-hashed UserID + BookID that never left the device, and only transmit the hash and the location information in a standardized format. This would give you anonymous page syncing between devices without having to even worry about encryption of the traffic, as long as you used an appropriate hash function. I would prefer this approach, because (as mentioned above), if the entire communications stack is encrypted, it’s a black box for anyone attempting to see inside and verify what the vendor is actually collecting. There are answers to this as well (encryption keys that the vendor never sees at all, for example, and are totally local to the user’s device a la Apple’s latest security enhancements).

There are technical solutions that satisfy our ethical concerns. We need to insist that our vendors care enough about our ethics that the technical answers become a market differentiator. We need to insist that this is important and then we need to make them listen.

By griffey

Jason Griffey is the Director of Strategic Initiatives at NISO, where he works to identify new areas of the information ecosystem where standards expertise is useful and needed. Prior to joining NISO in 2019, Jason ran his own technology consulting company for libraries, has been both an Affiliate at metaLAB and a Fellow and Affiliate at the Berkman Klein Center for Internet & Society at Harvard University, and was an academic librarian in roles ranging from reference and instruction to Head of IT at the University of TN at Chattanooga.

Jason has written extensively on technology and libraries, including multiple books and a series of full-periodical issues on technology topics, most recently AI & Machine Learning in Libraries and Library Spaces and Smart Buildings: Technology, Metrics, and Iterative Design from 2018. His newest book, co-authored with Jeffery Pomerantz, will be published by MIT Press in 2024.

He has spoken internationally on topics such as artificial intelligence & machine learning, the future of technology and libraries, decentralization and the Blockchain, privacy, copyright, and intellectual property. A full list of his publications and presentations can be found on his CV.
He is one of eight winners of the Knight Foundation News Challenge for Libraries for the Measure the Future project (, an open hardware project designed to provide actionable use metrics for library spaces. He is also the creator and director of The LibraryBox Project (, an open source portable digital file distribution system.

Jason can be stalked obsessively online, and spends his free time with his daughter Eliza, reading, obsessing over gadgets, and preparing for the inevitable zombie uprising.

8 replies on “Adobe Digital Editions and infoleaks”

ARRRRGH. I’m sorry, Liza. I’ve butchered your name twice. FIXING THAT NOW.

Thanks for the notes, Nate and Andromeda.

The local hashing of userid and titleid is a good idea. Worth trying to build a prototype to see if it would work in practice. But a fragment stream would likely have a fingerprint that could be used to crack the title unless you’re really careful.

Eric: you’d have to treat it like an actual password hash, so: use a good hash engine (bcrypt or the like) and use a salt. If done properly (always the hard part) there’s very little risk of exposing the title of the book in question.

/me *goes off to see if anyone is doing this*

Leave a Reply

Your email address will not be published. Required fields are marked *