ACM Copyright Licenses: Which should you choose, and how do you handle third-party material?

Choices, choices! This post should help you when you get to this part of the ACM copyright license.

Note that this was last updated in 2018, so may not account for changes in this form/process!

A few years ago, ACM changed its copyright model to offer three different options for authors publishing ACM papers: (1) a non-exclusive license to ACM that requires an “open access” fee; (2) a license granting ACM exclusive publishing rights; and (3) a copyright transfer to ACM. Previously, the third option was the only one available to authors, and this move was generally seen as a response by ACM to criticism over their lack of open access.

Since this change, when there is a new camera-ready manuscript deadline for a big conference, multiple people always ask my advice on handling their copyright. My typical tweet-sized response is: “License. No reason to transfer your copyright.” So I thought I would explain this in more detail, along with some related issues of what you’re allowed to do after you grant this license, and finally some detailed notes about something else I get a lot of questions about — use of third-party material.

This is based on some research and my reading of the copyright licenses, not from speaking with anyone at ACM — and of course, my opinion and IANYL. Most of the content in this piece comes from two blog posts I wrote a few years ago, one about licenses and one about fair use, though I’ve updated to reflect some changes in workflow that appear to be fairly recent.

The open access option
If you (or your research funding) care about open access and you can afford it, then option #1 (non-exclusive license) is great. This allows authors to grant ACM a “permission release” which is a non-exclusive license, and authors have the option to attach a Creative Commons license to it and to customize that license (i.e., noncommercial, no derivatives, and/or share alike). They can also attach a CC0 “license” to it, which essentially releases the work into the public domain. Regardless, this means that the paper will be openly available in the ACM digital library, and that you will also be free to distribute elsewhere.

Authors can still take this option while granting ACM an exclusive license or a copyright transfer, and this is included an an explicit option in the workflow once you indicate that you wish to pay the open access fee. However, I have no idea why anyone would choose to do this.

It is also worth noting that the open access cost has actually gone up substantially in the past few years; for CSCW 2018, the prices listed are $1,300 for ACM members and $1,700 for non-members. You can find more information about Creative Commons licenses and what they mean here. (Or ask me about it!)

the options for Creative Commons licenses that appear during the in the ACM permission workflow

The difference between license and transfer
If you don’t have $1,300 to spare, choosing between options 2 and 3 (license and transfer) will likely have very little practical difference. Transferring your copyright means that ACM completely owns that work. The other option is to grant ACM an exclusive license to publish your work, which, in practical terms, has pretty much the same outcome — the work is still exclusive to ACM. To me, it’s more a philosophical difference. The advantage stated by ACM to transferring your copyright is that it obligates them to defend against infringement. However, this is probably never going to come up, and even if it does, they have just as much interest in protecting the copyright either way. Transferring copyright will also put ACM in charge of clearing third-party rights for you, but that is also unlikely to be an issue for most of you. In my opinion, there is really no reason to entirely give up ownership of your work.

What does the license entail?
All that said, the license option is no great boon to open access since it’s no more open than it was before. (Though something that is useful for openness is Author-izer, explained in detail in the next section.) Here is the text of that license:

Owner hereby grants to ACM an exclusive, worldwide, royalty-free, perpetual, irrevocable, transferable and sublicenseable license to publish, reproduce and distribute all or any part of the Work in any and all forms of media, now or hereafter known, including in the above publication and in the ACM Digital Library, and to authorize third parties to do the same.

Part of my research has involved copyright licensing agreements, so I can tell you with some certainty that for most people this requires some translation. Here’s mine:

ACM has exclusive rights to this work, meaning that no one (including you) can publish it anywhere else, or give anyone else the right to publish/distribute/copy. There are no geographic constraints on ACM’s use. They do not have to pay you anything. This license does not expire. You cannot revoke this license if you change your mind. They have the right to publish this work in the ACM Digital Library or any other medium (including those that haven’t been invented yet). ACM can also license this work to others without further permission.

There is also a clause about “minor revisions,” noting that any derivative work that contains less than 25% new substantive material falls under this same license, and this new work must display the license and be labeled as a minor revision. (This of course brings up the related topic of self plagiarism, which is complicated because like many things, law, ethics, and policy don’t always align. Here’s an interesting CACM column by law prof Pamela Samuelson about it.)

How can I use my own work?
If you choose to pay the open access fee, then you are granting ACM a non-exclusive license to publish, which means that you can publish the work elsewhere however you choose. If you choose one of the other two options, then ACM provides some specific exceptions to the exclusive rights. As the author of the work, you (and only you) can:

1. Re-use any portion of the piece in future works you write or edit, such as books or presentations (and, I assume, dissertations, which is pretty important). This does not include contributing work to collections in which you are the editor, or to commercially produced course packs that are sold to student. You are also required to have a citation with a DOI pointer to the ACM digital library in any of these re-uses (including dissertations).
2. Create a “major revision” (defined as more than 25% new material) that you own (though it must cite the original work).
3. Publish the “accepted version” of the piece on (1) your personal web page, (2) your institutional repository, (3) any repository legally mandated by your funding agency, and/or (4) a non-commercial repository that does not duplicate an ACM table of contents or pattern of links. (More about this below!)
4. Use an author-izer link on your personal webpage or institutional repository.
5. Post the non-peer-reviewed (i.e., originally submitted) version to non-peer-review servers (which I assume means something like ArXiv or SSRN) prior to submission.
6. Distribute the final version to your employees, or for classroom or personal use.
7. Bundle the work with software distribution.

There are some obvious FAQs related to this list, and I’ve done a little bit of research on how people have interpreted it, but some of this is based on my assumptions.

• What is the “accepted version”? This means post-peer-review revisions, or what you actually prepare as opposed to any edits/preparations made by the publisher. This is actually a huge advantage for an author publishing with ACM because we prepare it ourselves — your “camera-ready” paper is virtually indistinguishable from what ACM publishes (I think the only difference may be page numbers). In short, you can put your camera-ready paper on your webpage. (Note that this likely means you technically shouldn’t post the final, nicely-laid-out versions of CACM or XRDS or other magazine articles.)
• Can I put the article on ResearchGate or similar? My answer to this question has actually changed since I originally wrote about this issue, in response to a 2016 update to the copyright license. I previously wondered if ResearchGate could count as a personal webpage, but I think that the license’s prohibition on commercial repositories makes their stance on this fairly clear. Only non-commercial repositories are allowed and are defined here as being “owned by non-profit organizations that do not charge a fee for accessing deposited articles and that do not sell advertising or otherwise profit from serving articles.” ResearchGate is for-profit, and so posting camera-ready versions there appears to be a violation of this license.
• What is author-izer? This is a pretty neat thing that ACM does that barely anyone seems to know about. Through your ACM author page, you can get a piece of html for each article that you put onto your webpage. It provides a link to the article that anyone can access for free. The advantages to this over just hosting the camera ready version yourself is that (a) it does provide the definitive, final version with any page numbers (or with nice formatting as in the case of my XRDS article), and (b) the download stats are added to your ACM page. Of course, this isn’t available until the piece is published in the database. All this said, I’ll add an honest caveat: I used authorizer for a while, but found it difficult to keep up with and the formatting clunky, so your mileage may vary.

If you were only interested in what copyright license to choose, and what it means, you can stop reading here! The next part of this discussion is specifically about what to do with respect to claiming third-party material when you are filling out your permissions form.

What about fair use of third party material?
In short, the process for claiming fair use of third party material for an ACM publication is confusing at best and misleading at worst, so this next part is my attempt to explain fair use and to give you an example of the way that I deal with this issue. Again, this is not based on conversations with anyone at ACM, not legal advice, and IANYL, but I hope that this information is helpful!

When you’re filling out the copyright form and have made your decisions about licenses as explained above, everything’s good until you get to this question:

the two choices available for third-party material

How do you know if you have “third-party material” in your paper? ACM provides some guidelines, and mentions things like “figures, tables, graphs, photographs, simulations, music or audio/video clips.” They also have a great section about using Creative Commons licensed material, in particular noting two important things: (1) ACM is not non-commercial, so you can’t use material licensed as CC-ND; and (2) You must attribute properly, which means more than just a link to the original source.

Copyrighted third-party material therefore most likely includes things like images from data that don’t belong to you (e.g., Instagram photos) or a figure or table referenced from someone else’s publication. The first time I dealt with this it was a paper I co-authored with Kurt Luther on his dissertation project Pipeline that included images of the cool creative things that study participants had made using Pipeline. More recently, a student working with myself and Shaun Kane had a poster at ASSETS (about crowdsourcing comic book transcriptions for blind fans) and the paper includes a comic book panel.

On this copyright form, you are given two options: (1) You don’t have third-party material; and (2) You have third-party material and have the necessary permissions. So basically, I have lied several times on this form. Because fair us is not a permission. Fair use is, in fact, about the absence of permission. The super TL;DR of fair use is that it is an exception to copyright law that lets you make some use of content whether the copyright owner likes it or not. This is a very important safety valve in the law to ensure that copyright cannot squash free speech. For a ton more information on this, see the Code of Best Practices in Fair Use for Scholarly Research in Communication out of American University. I also wrote a fair use analysis of my Barbie remix, or hey, I have a whole CSCW paper about it! (Side note: Fair use is, of course, U.S. law, though there are versions of this in other countries. It is relevant here, however, because ACM is using U.S. law.)

So whereas it should actually say “I have used third-party materials and have the necessary permissions and/or rights“, choose the second option anyway, and you will be presented with this popup window. (As a side note on “and/or rights,” that would also apply if the third-party content were, for example, public domain (like a photograph from NASA)).

the choices that appear when you indicate that your paper has third-party material

ACM provides guidance on how to fill this out, but not on how to claim fair use. It links to fair use guidelines, which are decent! I’m not sure I agree completely with all of their counter-examples (absent more context) but I think this is a pretty good document for thinking about whether something might be fair use in the context of a paper. (One thing I do appreciate about the guidelines is that it suggests that you consult your university librarian, as opposed to recommending you ask a lawyer. Good advice!) Though the problem is, there’s still no information about how to claim fair use in practice — which is why I’m writing this.

Unfortunately when it comes to the form itself, there’s even more fair use confusion. Clicking on the question marks gives you more information, so let’s go through each of these, and how you might want to answer them to claim fair use.

1. ACM Citation Reference (“Where in your ACM paper or presentation the third-party material appears… each third-party figure or image reference must be noted on the form and in publication.”) This is spot on. Remember that fair use means you don’t need permission, not that you don’t need to cite. You’ll see that on our ASSETS poster paper, the image caption includes “Image © DC Comics.” (We also included a cite in the references, because hey, it’s awesome to cite a comic, right?) So in this field you could write something like “Figure 1” or wherever the image + citation appears.
2. Original Third-party Source (“Where the third-party material was published or found”) Also spot on! This is where for that paper I gave the citation for the comic book. If yours is, for instance, an Instagram photo, then you could include the URL. If it’s a figure from someone else’s book, then a citation for that book.
3. Approved By (“Who approved your use in republication? Give copyright owner’s name (author or publisher) and include contact info. Note: Claims of Fair Use mostly apply to formal reviews and teaching, not to republication.”) And here is the problematic bit. Again, fair use requires no approval. Also, the characterization of fair use here is kind of strange. I don’t think that it is purposefully misleading, but it is almost certainly misleading in practice. After all, most people using this form would be using it for the types of examples I’ve indicated — not for wholesale copied/republished content. I’m reminded of the YouTube Copyright School video, where they basically say “yes fair use exists but it is so complicated you will never understand it, so you should just get a lawyer.” As for completing this field in my own experience, I’ve just written “This is fair use and does not require approval.”
4. Date Approved (“the date on which you received permission to use this content”) Again, no approval required for fair use, so I just put the current date.
5. Proof of Permission (“an email or pdf document granting you permission”) This is required for the form, and again is misleading because fair use does not require permission. However, this is the perfect place to include a fair use analysis in a separate document, which is what I do.

And here is where it’s worth pointing out that I have no idea what other people have been doing when it comes to this form. I’m just assuming that my practice of writing up a fair use analysis is not a thing that most ACM authors would think to or know how to do. So… is it the norm to just say that you have no third-party material even when you do? In which case if that’s been working all this time (i.e., apparently ACM doesn’t care) then it probably won’t hurt to keep doing that. It could be that if ACM were to revise their policies around this, they could end up being more arduous on authors, so there may be nothing wrong with the status quo.

But if you want to make sure you’re totally on the up-and-up, here is a copy of the fair use analysis I wrote and submitted with that form for the comics poster paper. (I have never gotten any response to these, so I can only assume that it works perfectly fine.) There are a number of links above that have more information about the fair use factors, though what I wrote there is a pretty good template for the type of things that might be relevant:

1. The purpose and character of the use. Criticism, education, transformation are key here. You can’t state that it’s noncommercial because even though you aren’t making money from it, ACM is (sadface). This is also a good place to state that the illustration is required to make your point from a research perspective.
2. Nature of the copyrighted work. Honestly even judges pretty much ignore this so I don’t worry about it. But it is favorable to fair use if the appropriated work is “non-fiction” rather than “fiction” (or if it’s unpublished).
3. Amount and substantiality. How much of the original did you use? The answer might be “all of it” but something that is relevant is the resolution of a photograph, for example. I usually note that I used only as much of the original as was required for my purposes.
4. Market harm. Is this material being in your paper going to harm the market for the original? Like is someone going to read your paper and then not purchase the original? This would often be irrelevant. For example, I noted that since we only used one panel from a comic, no one would read our paper instead of buying the comic. Or for an Instagram photo, since it was available online for free viewing, there was not really a market for the original anyway.

This was longwinded, but I hope it helps! I am a huge proponent of knowing and using your fair use rights. Unfortunately I have heard too many stories from academics about choosing not to include certain content in their papers because it might be copyright infringement. Chilling effects are bad!

I hope this was helpful! Please feel free to comment with questions, and though I can’t promise I’ll have the answers I can at least give you my thoughts. My advice would also always be that if there is something especially important/concerning, to contact ACM with specific questions about legal issues.