The Hazards of Mass Licensed Internet Digital Content For Film and Television Reuse

Tony Duarte (B.A., LL.B.) practices exclusively in the area of entertainment law and is an Adjunct Faculty member of Osgoode Hall Law School.

Perhaps one of the most problematic rights clearance transactions for a producer of film or television production is the licensing of existing photos, film/video clips, or music to the producer for reuse in its productions.   At best, the process is expensive and cumbersome for it typically involves the producer’s negotiation of multiple licenses one-to-one with a myriad of rights-holders with legitimate claims to rights in the same work.  For example, a single song may well have a number of copyright owners to be identified and located.  Thus terms of use of the song in the film or television production must be settled with each owner.  This approach is effectively dictated by copyright law, which recognizes the sole authority of the copyright owner to determine the instances and terms of licensed use of a copyright-protected work.  Many producers, therefore, become drawn to the relative ease and simplicity promised by internet stock sources of music, photos, and video content available for reuse in film or television productions and downloadable at the click of a mouse.  Upon closer inspection, however, many of these mass market internet alternatives are fraught with hidden transaction costs and legal risks of a different kind.

There are several principal kinds of internet stock content sources.  One type of website offers users content that has purportedly fallen into the public domain. Other websites offer users open access to their copyright-protected content under so-called “creative commons” or “copyleft” licenses.  Finally, some websites offer controlled access and use of their content on condition that users actively indicate their acceptance of what have been described as either “clickwrap” or “clickthrough” contracts.  These are essentially “contracts of adhesion” in that the terms of use are set out by the content-provider and are non-negotiable by the user.  Each of these internet sources provides its own hazards for any film or television producer intent upon using their offerings.

Websites offering works supposedly in the public domain often include notices disclaiming any representation or warranty by the site operator that the works presented are indeed outside copyright protection.  This would not be a serious obstacle to a producer’s reuse of such material in a film or television production if the authorship of each work and all other relevant facts regarding its copyright status were accurately given to permit the producer to evaluate the basis upon which the work is asserted to be in the public domain.  Unfortunately, unlike reputable public archives, many of these sites do not provide such information.  This leaves the prudent producer with little choice but to discard many of these sites as unreliable sources.

The websites which offer users content subject to “copyleft” licenses initially seem very attractive for the perceived fairness of their licensing terms.  The fact that some of these are owned by reputable non-profit groups with the laudable goal of promoting the free exchange of new works adds some legitimacy to their license terms.  Upon closer inspection of their licenses, however, there are at least a few issues that make many such licenses highly problematic for a commercial film or television producer wishing to use their content.

First, many of the copyleft licenses expressly provide that, in return for the user’s free use of the materials offered on the site, the user is deemed to agree that its own derivative work from such materials will, if released to the public, be made available to the public on the same copyleft license basis.  This aspect of copyleft licenses has been described as “viral”, or less charitably “cancerous”, in that, when using this content, some users may unknowingly put into the so-called “creative commons” all of the work they have added to the single, perhaps minimal, piece of copyleft source material they have used. 

Depending upon how one defines a “derivative work”, some copyleft licenses may be entirely incompatible with use of copyleft works within commercial film or television productions if such content is combined with other works that have been licensed by the producer under customary commercial licenses.  The usual terms of commercial licenses restrict the licensed use to the new production only.  That kind of contractual restriction is entirely at odds with a copyleft license which requires the entirety of the new production to be put into a copyleft regime – including all of the new production’s original material, copyleft material, and commercially licensed material combined.  Furthermore, most commercial productions are licensed by their producers for broadcast or distribution by licensees who typically contract that they are accorded exclusive exhibition rights, at least with respect to the original material of the production.  Works included in the production under copyleft licenses that apply the copyleft license to the entirety of the derivative work could put the producer in breach of these grants of exclusivity in the production.  Some copyleft licenses may also prohibit the derivative work from using any TPM (Technical Protection Measure), such as encryption in the DVD format.  Again, such license restrictions on the production as a derivative work may conflict with the right to use TPM measures licensed by the producer to the production’s DVD distributor.

Many copyleft licenses were initially written as computer software licenses.  Over the years, as the works to which these licenses have been applied has broadened to text and images, the drafting of these licenses has not been fully adapted.   For example, one sometimes encounters stills on the internet which are annotated as “copyright: GNU General Public License”, referring to the form created by the GNU Project and widely used.  The application of the GNU General Public License to images rather than software is a very awkward fit.  The license defines the work to which it applies as the “Program”, a term and definition clearly more comfortably applied to software than to images.  In addition, within the terminology of this license, how does one consider the resulting new television production which incorporates a single image that is governed by this license?   Is the production closest in description to a “work based on the Program, or the modifications to produce it from the Program”, in which case the new production must itself be subject to the GNU General Public License?  Or is the new work “a compilation of a covered work together with other separate and independent works”, which does not cause the license to apply to the other works?  Given the ambiguities that sometimes result from these ill-fitting definitions, the prudent producer often simply opts out of using any content that is subject to a copyleft license.

Copyleft also suffers from the proliferation of unauthorized variations to the original copyleft licenses initiated by reputable sources.  Since copyleft licensing is largely self-regulated, website operators can pick and choose which copyleft license to adopt and effectively write or edit their own.  As more unauthorized license variations come into existence, the initial value of copyleft licensing as an efficient and effective alternative to traditional one-to-one copyright licensing is accordingly diminished.  If copyleft licensing is not quickly gathered under the umbrella of a small number of authoritative sources of licensing terms with the resources to issue equitable and tailored terms for a broad range of user applications over a wide range of content, then copyleft content may never achieve widespread use in the commercial sphere.  In that respect, copyleft could be marginalized to the non-profit sector and fail to fully realize its potential to expand circulation of its content over all forms of copyright use.

The dissemination of unauthorized copyleft license variants can easily render the copyleft license alternative little different from the typical “clickwrap” or other adhesion contract option.  Contracts of adhesion, whether those used in car rentals or copyright, raise well-known issues: they are non-negotiable, often not read by the user, and invariably one-sided in at least some aspect in favour of the party who has drafted them.  In copyright licensing, they have exploded in number with the increased availability of content offered over the internet.  Some of these adhesion contracts demonstrate extraordinary examples of both incomprehensible drafting and one-sided terms.  For example, I have encountered at least one commercial internet image archive whose adhesion contract terms permit “royalty-free” use of their content for a small processing fee on condition that the entire production in which the downloaded image is to be included is deemed licensed by the user back to the internet archive for its own use at its absolute discretion.  This is a more disturbing variety of the copyleft’s “virus” because it raises all of the same alarming consequences for the unwary producer, yet cannot claim copyleft’s non-profit public policy rationale.    These types of contractual landmines, buried deep within the small print, taint the entire adhesion contract practice, seriously undermine any kind of efficiency that could be gained by licensing in this manner, and erode the confidence of producers in using many of these websites.

Copyright laws have favoured one-to-one licensing between each copyright owner and copyright user and have not kept pace with the mass licensing opportunities afforded by the internet.  In this respect, these laws have failed to serve both legitimate copyright users, such as film and television producers, as well as the copyright owners seeking to effectively harness the new possibilities of internet distribution.  Owners and users have turned increasingly to copyleft and adhesion contracts to find an efficient and effective alternative.  However, these alternatives have serious legal hazards for the prudent user which demand vigilance by their legal counsel and seriously undermine their intended commercial benefits to both licensor and licensee.  These hazards also make evident, nevertheless, that perhaps there is room for concerted government regulation to provide a process for stakeholders to arrive at more balanced, authoritative, appropriate, and wide-ranging standard form agreements that can make special purpose mass licensing of digital material more effective, efficient, and reliable to the benefit of both licensor and licensee.

5 Comments
  1. While it’s useful to highlight these pitfalls for the uninformed producer looking to use stock sources from the Internet, the post’s critique of “viral”, “cancerous”, or “share alike” terms ignores the intent of authors who choose to apply these terms to their works. In addition, there are some good solutions available to address some of the other issues raised without the need to resort to “government regulation”.

    When an author is choosing a particular copyleft licence, for example a flavour of creative commons licence, to apply to their work, they have a wide variety of options to choose from. These range from very permissive (only attribution is required) to less permissive (no derivative works or derivative works relicensed under the same terms, no commercial usage). That is, the author is able to make a choice as to how their works can be used by others. An author who only wants their work to be reused in similarly licensed derivative works can thus opt to do so. To then make an argument that the author is making it difficult for a film producer to reuse the work misses the whole point, the author specifically does not want it to be used that way!

    The exercising of this choice by authors can be seen on flickr’s creative commons page. Flickr users have decided to protect over 95 million photos with creative commons licences. The above link breaks down the use of licences based on which rights authors have opted to apply to their works. The variety in licences chosen reflects the fact that authors are making these choices.

    Copyleft licences are adopted by authors for a variety of reasons and it is wrong to make the assumption that all of them are to encourage the widest possible use of their works by others in all ways. Licences that include these “viral” or limiting terms do not therefore “seriously undermine their intended commercial benefits to both licensor and licensee”. They actually do the opposite by clearly expressing the licensor’s intended commercial usage.

    That being said, one of the good points raised in the post was that there is often no easy way to get in touch with an author to try to negotiate special or additional licensing terms. In 2007, creative commons began developing a protocol, called CC+, to provide “a simple way for users to get rights beyond the rights granted by a creative commons license”. Hopefully this new solution will spread quickly to continue to further expand the range of rights clearance options that an author can choose to make available.

  2. I take Ryan’s point that authors, or other copyright owners, may not want to allow unrestricted commercial or derivative uses of their content by others. In that regard, their choice of a copyleft license that restricts such uses may well be deliberate and should be given full effect by the user. Given the nature of these licenses as essentially contracts of adhesion, however, I am not convinced that as many users of these materials in derivative works or commercial applications necessarily appreciate or fully understand the import of these mass license restrictions.

    Although I am certain that many authors also invoke mass licenses without fully understanding their terms, it is more likely that the authors of posted works understand and deliberately choose mass licenses that reflect their intention since the choice of license is usually made by them. The user is more often simply attracted to the ease of access to this kind of licensed material and the cost-free or low-cost automatic license and is less likely to read the license fine print. The proliferation of copyleft licenses makes it even less practical for users to sift through each license variant and increases the likelihood of this kind of intuitive, rather than informed, understanding by users of the mass license used in each case.

    Time will tell if a small number of organizations such as Creative Commons will eventually become the standard and authoritative sources for copyleft, or perhaps also commercial mass licenses. I think that it can only advance the interests of both authors and users if that were to happen. The resulting licenses would be more likely to be properly understood by both parties and therefore more likely to be used as intended. Until that happens, however, users especially must be very cautious with regard to understanding each mass license for each work.

    In suggesting a role for government regulation, I do not see government stepping in to set the terms of such mass licenses or to determine who may create them. I see instead perhaps something as basic as government maintaining an official registry where an organization such as Creative Commons can register and deposit the definitive text of each kind of mass license it issues for public use. Authors can then invoke such a mass license by notice on the work that makes reference to the applicable license by its official registered designation. I think this would bolster the legitimacy, efficiency, and effectiveness of these mass licenses by reducing their number and source and by making it more likely that the user, and not only the author, would be aware of, and then actually abide by, the terms of use.

  3. I would add a couple other things to think about in connection with film production.

    Errors and Omissions Insurance: I don’t think I’ve seen an E&O carrier take a position on Creative Commons Corporation licenses one way or another. My expectation is that underwriters would probably view such licenses skeptically and coverage counsel (or more likely production counsel) would have to do some explaining or even better (for the carrier) put their own policy on the line. That skepticism will be lessened if the producer were to take CCC’s own advice and confirm the rights with the licensor, but that starts the iterative process that kind of begs the question about how meaningful the CCC license is in the first place.

    Cover songs/co-writes: A number of people have noted that the CCC “license” does not distinguish song from sound recording very clearly. This may be of particular concern with cover recordings or co-created works.

  4. Chris brings up an interesting point concerning producer’s errors and omissions insurance. Typically such policies exclude from coverage any claims that arise out of contractual disputes. Such claims are considered in the nature of a breach of contract rather than, for example, infringement of copyright, which is covered by such policies. Therefore, is every user that exceeds his or her creative commons “license” in effect excluded from coverage under errors and omissions insurance because the matter is merely a dispute as to the terms of a “contract”, or can it be argued that there really is no contract when it comes to mass licensing because the fundamentals of a contract are missing? While a “click-through” license may evidence offer and acceptance and support the existence of a contract, can the same be said if a work is merely offered up for reuse to others on a Creative Commons type of license in which the only evidence of “acceptance” is the reuse itself?

  5. Stepping outside a licence when using a copyright work may be both copyright infringement and breach of contract: see Jacobsen v. Katzer, 535 F.3d 1373 (Fed. Cir. 2008) on the open source licence. The point doesn’t just arise in insurance exception clauses. It cuts across things such as the Federal Court’s jurisdiction (infringement but not contract breach), limitation periods (different for the two causes) and calculation of damages (tort, including punitives, & statutory damages for infringement, not so for contract). I should be interested of any precedent that says differently from Jacobsen for an insurance exception clause. Given a contra proferentem construction, I am sceptical.

Comments are closed.