Find A Photographer Find An Assistant Join ASMP Join the Mailing List

When the Copyright Owner Can't Be Found...

In an article published in the mid-winter 2008 issue of Against the Grain, the newsletter of the American Library Association, ASMP Executive Director Eugene Mopsik and General Counsel Victor Perlman delve into the Orphan Works dilemma at one of its most consistently challenging points: securing permission to use photographs.


There is a real and on-going problem of what to do when the copyright owner cannot be located, or perhaps even identified. How does one proceed when it is impossible to obtain permission to use such a work or even know whether it is still protected by copyright? This dilemma merits a workable, effective and fair solution. The issue is popularly referred to as “orphan works,” but is really an “unlocated copyright owners” problem.


Legislation introduced in the House in the last Congress to deal with this conundrum had broad support, but it had one fatal flaw: it ignored the unique problems relating to visual artworks and would have savaged many of the copyright interests in photographs and other visual artworks. To vastly oversimplify a fairly complex piece of legislation, the basic approach was that a prospective user of a copyrighted work had to make a reasonably diligent search to locate the copyright owner. If that search were unsuccessful, the user would be free to use the work. If the copyright owner later came forward, the user could stop the use promptly and pay no compensation. If the user could not or would not promptly cease the use, his only obligation would be to pay “reasonable compensation” for the use. In neither event would the user be exposed to the possibility of having to pay statutory damages or the copyright owner’s legal fees, which would normally be a possibility in a traditional copyright infringement situation. This issue, and a similar approach, is likely to arise again in the current Congress. Fortunately, there may be a simple solution that we will explore later in this article.


To understand the problem, one must first understand the customs and practices of the trade in publishing visual images, the natures of search technology, and the procedures of the U.S. Copyright Office. A fundamental fact is that most images are published without attribution attached to the image. In the print world, magazines and similar publications account for a huge inventory of photographs and other visual artworks. A large portion of those images appears in advertisements. Advertisements virtually never provide any credit or by-line for the photographer or artist. Editorial uses sometimes appear with credits next to the images, but most often provide no attribution or place the credit in small print at the bottom of the page or the back of the publication, or provide no credit whatsoever. Exacerbating that situation is the fact that all it takes is a pair of fingers to remove an image from its context, including attribution, completely. Then, add to that the untold millions of photographs tucked away in drawers, shoeboxes, attics and albums all over the world. Most of them have no attribution on them and become virtually untraceable once they leave the hands of their original owners — assuming even their original owners know who created them or who owns the copyrights to them.


The situation in the virtual world is even worse. Most Websites provide no credit or attribution whatsoever. Multiplying the problem exponentially is the fact that, according to industry service providers, as much as 90% of the visual images appearing on the Web are posted without authorization or knowledge of the copyright owner. Further, identifying metadata embedded into an image can be stripped out inadvertently through the most mundane of image manipulations, such as changing image size or resolution.


Because of these and other factors, the bottom line is that most of the visual images that exist are in forms and locations that provide no direct attribution, or even contextual information that could be used to identify and/or locate the copyright owners. Even when there is contextual information, such as the identity of a publisher or advertiser, those parties have neither the resources nor the motivation to do the necessary research to respond to inquiries by third parties. In fact, under legislation like the bill introduced in the last Congress, professional users of images, like publishers and advertisers, would have a vested interest in helping to create and expand the pool of “orphan work” images available for use at little or no cost.


Given all of this, traditional “analog” search measures are extremely time-consuming and generally unlikely to bear fruit. One would think that the wonders of digital technology and the Internet with its massive search engines would resolve the issue. In the context of finding information about textual materials, that is generally true. However, almost all of the search technology is based on text searches — and text searches simply do not work on visual images. One could attempt a search by keying in words that describe the contents of the image in question. However, that would not necessarily yield the image or any information about it. If the image were in a database with matching keywords attached, the search would produce a hit. Unfortunately, even in that successful case, the search would probably be too successful. Consider how many times you have seen pictures with a beach, blue water, a clear sky and palm trees. Putting those words into a search would yield a universe of hits, none of which is likely to be the one that the searcher is seeking.


Making the situation worse is the fact that, as mentioned above, even if identifying metadata is embedded in the image, it is likely to have been removed, either inadvertently or intentionally. Text-based searching is not a solution.


Recently, several companies have developed workable image-based search technology. A searcher provides a digital file containing an image. The vendor digitally analyzes the image to create a digital “thumbprint.” It then sends out a spider to crawl the Web in search of images with matching thumbprints. When it finds matches, it reports them to the searcher, who can visit the Websites where the matches appear and take appropriate action. The technology was created to help copyright owners locate infringements of their photographs, but it can readily be used for the reverse purpose of working from a copy to find the original, and its owner.


This technology is in its early stages of development and commercial use, but it is viable even at this point. It would appear to be panacea for both owners and users of visual artworks. That appearance, unfortunately, is illusory. For it to work, the image must be in digital form in a publicly accessible location on the Worldwide Web. If the image is not on the Web, the search engine cannot locate it. While there are many images on the Internet, and even assuming that a substantial portion of them are posted in a way or location that would allow a searcher to identify and locate the copyright owner, the vast majority of existing visual images does not exist in digital form, let alone on the Internet.


Copyright protection goes back to 1923 for published works, and even further for unpublished works. Only a small portion of the visual artworks still protected by copyright was created in digital form. Probably an equally small portion of analog images still under copyright has been digitized. And only a fraction of all of those digital images is on the Internet. For an image-recognition search engine to work effectively, it has to have an adequate universe of images to search. That universe simply does not exist on the Web, and it is likely not to exist for a very long time, if ever.


Some representatives of the U.S. Copyright Office and the publishing community have said that it is the responsibility of the creative community to digitize its archive of analog images and deposit them in private databases accessible on the Internet. In theory, that position seems to make sense. However, it ignores some crucial realities. First, we are talking about fundamental changes in the Copyright Act upon which business models have been in place for decades. As a practical matter, the legislation that was introduced last year would have the effect of making visual artists retroactively lose much of their previous copyright protection for any of their images that are not posted on the Internet with robust identifying information. There is a certain basic unfairness about that approach.


Second, and perhaps more significant, is the fact that most artists, illustrators and photographers are sole practitioners or tiny enterprises. They have little or no support staffs. They work full time to earn an income that averages something in the area of $30,000 - $40,000 per year. Each of them has an archive of analog images that probably numbers in the hundreds of thousands. For them, having to edit, digitize, and keyword their existing archives of analog images would mean having to pay someone else money that they don’t have, or stop working for a long period of time to accomplish those tasks, themselves. That is, the choice is to protect their existing copyrights by ceasing to work, or abandon those copyrights in order to continue making a modest living. This is, at best, a Hobson’s choice. The cost of a project of this magnitude is clearly demonstrated by considering the resources of the two entities that have already attempted it on a limited scale: Corbis, owned by Bill Gates, and Getty Images, started by the Getty family and publicly traded.


Probably the most logical candidates for creating a digitally searchable archive of images currently in analog form are the U.S. Copyright Office and its parent, the Library of Congress. They already possess a substantial archive of images through deposit copies of registered and/or published works. They also have access to the resources of the U.S. government, at least to the extent that Congress will allow. Since helping users to locate copyright owners is a public service, it is a task that should be well within the charges of both entities. However, the Copyright Office has taken the position that it will not, under any circumstance, allow online searches of deposit copies, even after it has completed implementation of its online system for registering copyrights and depositing copies. That is a Gordian Knot that appears to be too strong for any sword to sever.


Where does all of this leave us? The goal here is to provide a solution that will allow users to have access to copyrighted works in spite of the inability to locate the copyright owners while not sacrificing the rights of photographers, artists and illustrators on the altar of orphan works. One elegantly simple solution would be to take the previous legislation and limit the scope of the protection afforded under it to non-commercial uses that are non-revenue producing or are works of non-fiction. This would leave intact copyright protection where uses are for things like advertising, product packaging, t-shirts, coffee mugs, etc. However, it would provide a safe harbor for uses such as non-fiction books, articles, special interest Websites, documentary films, etc. As always, the devil is in the details, but we believe that this concept paves the way to a fair and workable solution to this quandary.


Currently, social networks and other file sharing Websites have created tremendous pressure to make images free and accessible to all. This makes the ability to identify a rights holder and track the various uses made of a particular image of paramount importance. Growing public pressure may, over time, lead to a diminution of the copyright protections currently afforded to creators. It is the interests of the public that drive legislation, not the concerns of rights holders. We need an equitable solution to the Orphan Works dilemma — particularly for images — and the maintenance of copyright protection in order to ensure the viability of photography as a profession.


This article first appeared in a publication of the American Library Association, Against the Grain Volume 19, No. 6 — December 2007-January 2008.