May 26, 2024

DMCA Week: A second orphan works problem?

The orphan works problem in copyright is real and serious. Several congressional hearings and a Copyright Office inquiry that drew hundreds of thoughtful comments—not to mention countless articles and blog posts—attest to that fact. This attention is heartening, and while orphan works legislation seems to have died this year, I’m optimistic that the next Congress will address the issue. As is often the case in Washington, however, such a victory might only mark the beginning of the next battle. The way I see it, the DMCA might create a second orphan works problem.

As you may know, an orphan work is a work under copyright the owner of which cannot be found. For example, say you come across a self-published political manifesto from 1967 in a Berkley archive or garage sale. You’d like to excerpt extensively from it in a book you’re writing about the Summer of Love. You try every possible avenue to locate the pamphlet’s author to get persmission, but you fail. That manifesto is an orphan work: it’s under copyright, but you can’t find the copyright’s owner.

The problem with orphan works is that if you nevertheless use the work without getting permission from the owner, you expose yourself to an infringement lawsuit if the owner later appears. Because statutory damages can run as high as $150,000 per infringing use, most orphan works go unused. This is a loss not only to the potential user, but also to society at large because it will be deprived of the promotion of science that would have resulted from a derivative work. Perhaps worse, an orphan work might be lost altogether because making an archival copy—say from fragile film to a more stable digital format—can be considered an infringement.

I have previously proposed a solution to the orphan works problem that would create an orphan works affirmative defense to infringement similar to fair use. Under this scheme, if you could show that you took every reasonable step to find a copyright owner and came up empty, you would not be liable for infringement. The Copyright Office made a similar recommendation, but instead of serving as a defense, showing a reasonable search for the copyright holder would merely limit the possible penalties for infringement. A bill based on that recommendation passed the Senate in September but never got a vote in the House before it adjourned earlier this month.

So what does this all have to do with the DMCA? My concern is this: Even if a strong orphan works bill were to pass Congress so that one would no longer have to worry about liability for copyright infringement, the work might still be unusable if in order to gain access to it one had to circumvent a technological measure in violation of the DMCA.

This is not a far-fetched idea. The Internet Archive has already successfully argued for a DMCA exemption for “Computer programs and video games distributed in formats that have become obsolete and that require the original media or hardware as a condition of access, when circumvention is accomplished for the purpose of preservation or archival reproduction of published digital works by a library or archive.” It needed this exemption to legally preserve legacy software stored on fragile floppy discs. Without that exemption, it would violate the DMCA even if it did not violate copyright.

You can see this problem presenting itself again. For example, the Prelinger Archive serves to collect and preserve ephemeral films of historical significance. According to its web site, “Included are films produced by and for many hundreds of important US corporations, nonprofit organizations, trade associations, community and interest groups, and educational institutions.” Today its collection is composed largely of videotapes and films, but there will come a time when one-of-a-kind movies will be on CSS-protected DVDs. Similarly, Amazon recently introduced its Digital Text Platform, which allows users to self-publish books that can be purchased and read on the Kindle. This means that there will soon be many books that will exist only as DRM-protected e-books. Therefore, even if we address the orphan works problem so that a user must no longer fear a surprise infringement suit from a previously impossible-to-find copyright holder, the user might still fear a DMCA suit.

The triennial exemption process provided in the DMCA will likely not provide sufficient relief because the Copyright Office is limited to exempting particular “class[es] of copyrighted works.” Just as it has refused to exempt “fair use works” because that is not a “sufficient” or “cognizable” class, the Copyright Office probably won’t recognize orphan works as a class that can be exempt. The sort of classes it will recognize will be very narrow, such as the one in the “obsolete video game or software” exemption. Not only is this exemption for one particular type of work, but it only applies to circumventions made for archival purposes.

Additionally, as Tim Lee points out to me, another way the DMCA might exacerbate the orphan works problem is by preventing the conversion of works into open and widely supported formats—the digital equivalent of what Prelinger is trying to do with film. Most of the proprietary DRMed formats we see around us today are likely to drop out of commercial use within the next couple of decades. As a result, people will gradually forget how to read those formats at all. By 2108, even if the DMCA has been reformed, no one may have any clue how to decrypt a PlaysForSure-encrypted audio file from 2002. Digital libraries in the near future need to be able to say “Boy, this format isn’t commercially supported any more, we’re going to convert it to MP3/MPEG/PDF so our patrons can continue enjoying it.” If they’re not allowed to do that, DMCA reform in the distant future may not matter.

I’m afraid I don’t have a ready solution short of abolishing or limiting the DMCA. One approach might be to include a limit to DMCA liability in the proposed orphan works legislation. However, I wouldn’t want to endanger that legislation’s political viability to address what is still a speculative problem. It won’t be long, however, before we find out if DMCA protections cause a second orphan works problem, metastasizing the harm visited on culture and society by that regrettable law.


  1. Carlie J. Coats, Jr., Ph.D. says

    My Ph.D. is in Mathematics (MIT, 1978). I have studied under some of the top logicians of the Twentieth Century. I claim that my opinion on matters of logic is expert; in fact, far more expert than many in the legal community.

    It is the plain language of the constitutional requirement, that its protection extend only “for a limited time.” The DMCA does not meet this test, and as far as I am concerned, it is unConstitutional on its face.

    That the matter comes up at all is a serious indictment of the President, the Congress, and the Courts who are forsworn in their duty to the law.

  2. Mission Possible says

    I don’t understand why so many art organizations are fighting against the Orphan Works bill when they are too cowardly to speak out against people who are already violating the copyright of other artists as Shepard Fairey has done. He is a prime example of the type of artist we should be going against instead of supporting. Put up or shut up. Look up what he did with the art of Rene Mederos. Or are you afraid of making the Obama crowd angry? You can’t have both worlds you know! We can’t stand for some rights some of the time. We must stand against ANYONE who steps on a fellow artists rights. The fact that he is so popular shows just how much the public cares for our rights!

  3. Your post addresses the case in which a work’s author is not traceable, leading to an inability to secure an appropriate license.

    Another situation that we’ve begun to see as entire DRM schemes have become orphaned is the situation in which a person who has complied with the DRM regime, licensing works in good faith and paying for the right to use them. When the distributor fails or decides to abandon the DRM scheme, the unfortunate licensee may find himself losing access to the works that he had paid for legitimate access to. DMCA seems to provide no protection for the legitimate licensee for this sort of passive taking, which seems rather unreasonable. As someone noted recently, a pirated work won’t suffer that fate, but one that is orphaned by the distributor will.

  4. The EFF’s published document, DMCA – Ten years of unintended consequences, cites many occurrences in which the DMCA’s acts and provisions wrongfully stifled competition and research in the technology sectors. I was hoping you could enlighten me as to your thoughts on how the DMCA has affected these issues, competition and research, and whether the EFF’s claim is good or bad? Why or why not?

  5. The problem of orphan works and loss-by-DRM actually arise only because the period for which copyright is granted is too long, without intermediate update. If copyright lasted 10years, renewable every 10 to 50, and was coupled with a compulsory upkeep requirement, where the work is DRM’ed ( contra, the loss of DMCA anti-DRM rights) the balance would be better between both sides. If the owner wants to cease offering DRM work-around access, then the DRM clause ceases to restrict non-infringing uses of the work. And if the owner has to re-license every 10 years, there would be fewer ‘unknown’ situations. If the owner does not care to assert his copyright, then he loses it. Which of course was the original intent of copyright law, before copyrights became just another legal formula to ensure collection of dane-geld.

  6. Clive Robinson says

    The are two main problems with an orphan work, who it actually belongs to and if they still have an interest in it.

    I’ll start with the easy one,

    Many centures ago it was common practice that after a period of time property was forfit in one way or another if the previous owner did not renew their claim.

    This principle is still alive and well in various ways in law. Perhaps the most obvious being the seven year rule before assumption of death is confirmed. That is if you abandond your goods chatels and contact with those who who might have an interest in them then the interested parties could have you declaerd dead and your title to the property vacated.

    I am quite aware that this has significant downsides, however it can be viewed as a place to start.

    In the UK there used to be a requirment that published works be lodged with the British Library ostensably for “the common good” in that it maintained a(n) historic record.

    Such a system is realy a must even in a modern society.

    Perhaps a small edition to the system where the owner of a work is required to re register their interest in the work would be a usefull addition.

    Additionaly a method by which the owner might be “officialy” contacted by the register if somebody expressed a wished to the register to make use of a work.

    Which brings me around to the second more difficult problem “who actually owns the work”.

    It has been said only half jokingly that “an authors name is not his own, it is owned by his public, publisher and agent”. Likewise the name on a work might not actually be the owner of a (ghosts) work. That is for some reason the moral owner of the work has either chosen not to use their own name or has been by way of trade given up part or all of the right of association with the work.

    Further in music it is often seen that a tune or song or arangment is marked as been “traditional”.

    That is the owner of the work has supposadly lost to the mists of time, but the work due to it’s popularity or importance has not.

    There are three points here ownership of the “original work”, ownership of a “derivative work” and what differentiates an original work from a derived work in ways that are acceptable.

    The big problem is that like patents original work is usally not of great benifit to the moral owner it is the secondry works that are.

    That is in a song might be very popular but not have sold well (or been offered) as sheet music originaly but at a later time a collection of like pieces in which it is included sells well (Eric Clapton amongst others had this problem). Likewise with the words to a song.

    So you realy have to ask not only,

    “Who owns the work both moraly and by way of trade”


    “Who has an interest in it via derived or later usage”

    And only then

    “Who should benifit from it”.

    Establishing this for a work in it’s original and derived forms is a gordian knot of considerable size and complexity, and unfortunatly as you have found IP agents and their legal helpers are busy adding fresh tangle to the mess…

    Hopefully a traceable register of interests might help solve the problem but patents and actions by IP holders sugests not 8(

  7. […] Perhaps worse, an orphan work might be lost altogether because making an archival copy—say from fragile film to a more stable digital format—can be considered an infringement. […]

    In such cases, the only morally right thing to do is to violate copyright law in an act of civil disobedience. It is better to risk litigation than to risk the irreversible loss of cultural artifacts.

  8. It could get even worse — once a device becomes obsolete, the very fact that some set of content is built around that device becomes an effective access control mechanism.

    (The orphan works bill that died in the last session was imo far too skewed to the interests of content republishers — too easy to declare a work orphaned, almost impossible for the original author to gain redress once a work has, rightly or wrongly, reached orphan status. I hope future iterations will strike a better balance.)

  9. How about if the copyright office defined a class which consisted of DRM’d media for which the publisher of the DRM system has withdrawn support. In theory, the only way to access this material would be through a DMCA violation in any case.

    In this class, use could be permitted for all fair use cases, or more likely it could be limited to archiving, non-commercial personal use, and scholarly publications.

    If a DRM purveyor doesn’t wish the works that they sold to fall into this class, then they have to continue to support the DRM that they sold to their victims (err. customers). If a copyright holder doesn’t wish their works to fall into this class, then they need to obtain contractual assurances from the DRM provider that the DRM will be available for the duration of the copyright (or they will get some other form of compensation).

    Just a thought!!!

  10. Don’t forget the problem of DRM that “calls home” as part of the decryption process. When the original distributor shuts off the “home” servers, the only way to access any of that material (all legally purchased) is to crack the DRM and thereby violate the DMCA.

  11. Frater Plotter says

    The other day it occurred to me how to turn artists and authors against DRM: make it clear to them how DRM defines the economic role of their work. Specifically, point out to them that DRM is only good for throwaway works.

    Is your work intended to be ephemeral? Is it something to be bought, consumed, used up, and thrown out a month later? Is it your intent to be replaceable and forgettable — indeed, do you want a business model that depends on your work being forgotten and replaced? Is your art 100% commercial and 0% personal? Does it contain nothing whatsoever that you wish to be remembered for? If so, then DRM is the solution for you!

    DRM means that your work will not live longer than the media technology upon which it is released. It means that your readers and listeners will inevitably lose access to the copies of your work they have “bought” — within a matter of months or years.

    Here’s how DRM makes your work ephemeral and replaceable:

    Technological devices such as iPods and Kindles have a limited lifetime, usually expressed as Mean Time Before Failure (MTBF). Any copy of your work that is bound to a specific device cannot last longer than the device itself. Perhaps the user will be able to get a new copy (see below) but perhaps they just won’t bother.
    Many DRM schemes rely on authorization servers to let users get new copies if their computers or devices fail. When your label or distributor decides to shut down remote-authorization servers, readers or listeners lose access to your work. Even if the publisher doesn’t want to shut the servers down, accidents happen, as do business failures. Your work will not last any longer than your publisher’s business.
    Technological devices and formats become obsolete. In fifty years, the iPod will have gone the way of the gramophone: a well-loved classic of a time long past. Even if that iPod itself has not broken down, users won’t be able to get at the contents of it, since they won’t have computers with USB connectors. So if your work is bound up on iPods and can’t be copied off, it will be lost along with the ability to access it.
    We can read ancient literary works because they were copied by generations of scribes, printers, and translators over hundreds or thousands of years. DRM and the DMCA make this ancient scholarly activity illegal, so your work will certainly not be legally available to your distant descendants. If they have it at all, it will be by the good graces of pirates.
    Some DRM systems allow for remote disabling of users’ access to a work that they have “bought”. This can serve as a form of automatic censorship. If a shifting political climate makes your work more controversial, the publisher or distributor has the ability to destroy the copies of your work that are already in users’ possession. In a world where DJs were forbidden to play ’60s peace songs in the weeks following 9/11, your work may be suppressed years in the future … and with DRM, that suppression can be technologically enforced.