Copyright Scholars Urge Reversal in SAS v. WPL

Fundamental copyright doctrines and procedures are presently on trial in the case of SAS Institute v. World Programming Limited, now on appeal at the Federal Circuit. Suffice to say, U.S. software developer SAS alleges copyright infringement by UK developer WPL, and these entities have been litigating on both sides of the pond for many years. But it almost doesn’t matter, for the purposes of this post, which entity ultimately deserves to prevail on the actual merits. Because according to a brief filed by ten copyright scholars, the district court in the Fifth Circuit bollixed up procedure and doctrine so badly that, if allowed to stand, the ruling would be harmful to copyright creators everywhere. And, as usual, the precedent would be highly prejudicial to small business and independent authors who are always at a disadvantage vis-à-vis the cost of enforcement of their rights.

At issue is one matter of settled copyright doctrine and one matter of settled court procedure, both of which the SAS court in Texas undermined with an unprecedented procedure of its own that it called a “Copyrightability Hearing,” which the scholars’ brief describes as “foreign to copyright law and threatens to impose additional costs and burdens on parties, moving the costs of vindicating their rights further out of reach for many creators and copyright owners.”

Doctrine: Unprotectable Elements Do Not Deprive Whole Works of Protection

It is axiomatic that nearly all copyrightable works contain unprotectable elements. For instance, the scholars’ brief cites nonfiction books as an archetypal subject matter that is unquestionably granted protection as whole works, despite comprising many facts, which are unprotectable. But to further illustrate the point, I would note that the average motion picture or TV series contains hundreds, if not thousands, of protectable elements intermingled with unprotectable elements.

In fact, the motion picture industry thrives on copyright’s limiting doctrines like the idea/expression dichotomy, merger, short phrases, and scènes-à-faire. Consequently, no filmmaker gets an exclusive to produce scenes depicting heroes running with guns or that first kiss as the music swells and the camera adjusts to reveal a romantic setting. But these limitations do not obviate copyright attaching to each motion picture as a whole work, or even segments of each film as original expressions of common ideas, themes, tropes, etc.

Now, imagine a trial in which the filmmaker alleges copyright infringement of a whole movie, or a substantial portion of the movie, but the defendant argues, “Well, the motion picture is full of unprotectable elements and is, therefore, deprived of all protection as a complete work.” Because that is essentially what WPL argued in its lawsuit with SAS, and the district court agreed, thus, upending doctrine by voiding copyright in the entire work. As the amici state:

The Court’s stated basis for rebutting the presumption of validity and holding the SAS copyrights invalid, i.e., that “at least some of the material is not entitled to protection,” is not only wrong, it is exactly the opposite of the correct standard. [Citation omitted]

I use the motion picture example to emphasize the point that if the only response required of a defendant were to recite a litany (as WPL apparently did in this case) of copyright doctrines  under which elements of a work may be unprotected,[1] much as I did a couple of paragraphs ago, in order to shift the burden to the plaintiff to re-prove protectability, that way lies madness. For instance, I happened to watch the new Suicide Squad this weekend and, even as a non-attorney, I could probably do ten pages or so breaking down the protectable and unprotectable elements just in Harley Quinn’s solo fight sequence.

But hypothetically placing the burden on the film’s producers to engage in such an exercise would be anathema to more than a century of legal doctrine and decades of court procedure. Yet, that is precisely what the court did in SAS when it found that, in general, the work at issue contained unprotectable elements and, therefore, the plaintiff bears the burden to re-prove copyrightability in that which is allegedly protectable.

Procedure:  Registration as Proof of a Valid Copyright

So, it is settled law that copyrightable works may comprise even large volumes of non-protectable elements and still meet the standard of originality for protection. Moreover, once a copyright application is approved, a certificate of registration issued by the Copyright Office is considered prima facie evidence of a valid copyright as a matter of law. In fact, the efficiency of this longstanding precedent is one of the key incentives to authors provided by Congress (in §410(c) of the copyright act) to register their works with the USCO. Here, the scholars’ brief argues that by shifting the burden to the plaintiff to re-prove copyright in its work, the district court has disturbed a presumption of copyright validity, stating:

This presumption is of vital importance to copyright owners, and is frequently cited as an inducement to register works although copyright protection attaches automatically upon fixation. The presumption is particularly important to individual authors and small entities who struggle to afford the costs of federal litigation to adjudicate copyright claims. Imposing the costs of a Markman-like evidentiary showing of originality in every litigation would effectively render even their timely-registered works judicially unenforceable.[2]

The Abstraction-Filtration-Comparison Analysis

A court may certainly write an opinion finding that a registered copyright in a work to be invalid, but not without thorough analysis and clear articulation of its reasoning. Here, the scholars’ brief argues that the district court compounded its errors by engaging in an incomplete and improper analysis of the work at issue and then exacerbated harm, not only shifting the burden to the plaintiff to re-prove copyrightability, but to do so within the context of this court’s erroneous analysis.

Not unlike the analyses for “substantial similarity,” the Abstraction-Filtration-Comparison (AFC) test is a judicial review which examines the work by filtering out the unprotectable elements and then comparing the remaining protectable elements to the allegedly infringing work. But here, the scholars’ brief notes examples whereby the court misconstrued features of the SAS work as a foundation for finding non-protection. One example cited is “open source elements,” which tends to confuse a lot of people because “open source” generally refers to software that is made freely available, but this attribute does not alter the copyrightability of the work. Thus, the scholars’ brief states:

…the Court’s attack on the protectability of SAS works is predicated on an alleged characteristic of the works that has nothing to do with copyrightability. Thus, by its own terms, the District Court filtered out what should be presumed to be copyrightable elements of Plaintiff’s work. This alone is reversable error.

As stated above, whether SAS or WPL should ultimately prevail on the true merits of each claim is immaterial to the issues raised in the scholars’ brief. And to be honest, I am not about to crawl through the thicket of facts, allegations, and responses presented in what looks like more than a decade of these two companies duking it out with one another. Some have referred to this case as a Google v. Oracle light, but at least in context to the brief discussed in this post, I would disagree.

In Google, the core copyrightability question (which the Supreme Court failed to adequately answer) was solely based on the merger doctrine, and whichever view one held of that argument, either could find purchase in the law. In SAS, it appears that the district court simply abandoned any foundation in law whatsoever. And for the sake of rightsholders working in all media, the Federal Circuit must reverse.


[1] The amici scholars refer to this litany as “the List” in their brief.

[2] Markman refers to judicial interpretations of specific words used in a patent description pursuant to a patent litigation.

Also see brief co-author Steven Tepp’s piece at IPWatchdog.

Photo by: AndreyPopov

Is Instagram (Facebook) Blocking DMCA Takedowns?

When the news broke that Charlie Watts had passed away, Instagram, Facebook, et al naturally bloomed with tributes, editorials, eulogies, and personal notes of gratitude for the late percussionist’s contributions to music. But although the virtual vigil has become standard practice every time a beloved cultural figure passes away, one overlooked difficulty of this and similar trends is that all that sharing is not particularly good for the photographers. Those men and women you likely never heard of who captured the images tend to lose opportunities as a result of all that sharing.

The force that motivates fans to search for that iconic photo of Watts or Ed Asner and add it to their social feeds is, of course, timeliness. But timeliness is also what drives news media and other commercial enterprises to license photos that accompany obituaries, tributes, creative commentary, etc. So, the problem with all that sharing (loving though it may be) is that it dilutes the value of these images, and often during the same periods when interest in the images is at its peak. And that does not even account for the commercial entities that publicly display images without license “because it’s already all over the internet anyway.”

As Martin McNeil, a former pro photographer in Glasgow tells me during an interview, “The only way to preserve the value of your work, truly, is to prevent dilution.” And that is precisely why he sent thirty-four DMCA takedown notices to Instagram directed at pages displaying his photograph of Ray Harryhausen, all uploaded around the May anniversary of the filmmaker’s passing (in 2013). The photo depicting Harryhausen with several of his iconic creature sculptures, was captured at the Edinburgh International Film Festival in June of 2008, and McNeil says he is used to seeing it trend on social media in May and June, corresponding to the observance of Harryhausen’s passing and birthday, respectively. But this year was the first time McNeil says that has ever seen Instagram erect a gauntlet of roadblocks to his takedown requests.

Instagram Exceeding Its Authority Under DMCA

I mention the dilution problem in order to frame the broader economic challenge for visual artists, but the plain fact is that the takedown provision of the DMCA does not require a stated reason for a copyright owner to send a request to an ISP to remove a work. As long as the sender of the notice is the legal owner of the copyright, or an agent of the owner, and the rest of the notice conditions are met, the default terms of the statute require that the ISP “expeditiously remove” the material. More information may be required from the copyright owner if the user/uploader files a counter-notice alleging that the takedown was wrongful (either for administrative or legal reasons), but no questions of that nature apply to the incoherent and legally untenable responses that McNeil received from Instagram regarding his takedown requests.

As McNeil details in a thread comprising over sixty tweets he addressed to Head of Instagram Adam Mosseri, one of his takedowns was honored right away, while the other thirty-three triggered boilerplate emails asking for “An explanation of how you believe the reported content is infringing your copyright.” “I was knocked back on my heels,” McNeil tells me, and for good reason. Because Instagram has no business asking this question. As such, he responded to all thirty-three messages with his own boilerplate, correctly stating that nowhere in the §512 of statute is there a requirement for the claimant to follow up on a properly filed takedown notice with further legal explanation as to why a use is infringing. The fact that Instagram asked the question at all is anathema to the foundation of §512, which was written to immunize platforms as neutral providers that bear no liability for the actions of their users.

After his follow-up responses to the thirty-three emails, McNeil received partial satisfaction, but in a manner that is even more absurd than the first round of responses. Because with regard to eleven his requests, the Insta-minions told McNeil that “…it’s not clear that you have rights to the reported content. In particular, it does not appear that you (or your client) created the reported content, or that you are otherwise authorized to submit this report.” Um. What?

Instagram has no authority under the terms of the statute to allege that a notice sender is not the owner of a work, or agent of the owner, subject to a takedown request. All DMCA forms, by law, contain a statement that, under penalty of perjury, the filer warrants that to the best of his knowledge, that the information provided is accurate. And that is supposed to be the end of the conversation for the ISP. Only the user/uploader of the material is empowered to respond if he believes, in good faith, that the notice filer has misrepresented himself as the copyright owner or agent of same. Remember, the ISP is allegedly neutral under the law, which is why it is shielded from liability in the first place.

So, to add outrage to absurdity, the fact that twenty-two Insta-minions removed McNeil’s photo as requested, while eleven of their Insta-colleagues accused McNeil of lying about his ownership of the same photo, should make any reasonable person ask the obvious question:  WTF is going on at Instagram?  Is this about inconsistency in training among the staff assigned to address DMCA matters? Or is there something special about the eleven uses that Instagram is reluctant to remove? Or is there fresh policy at Facebook/Instagram to slow walk takedown requests and/or construct new obstacles to dissuade rightsholders from enforcement?

After all, the average photographer, especially the one just starting her career, is likely not as legally savvy as Martin, who happens to be pursuing a law degree today. And from Facebook’s point of view, every valid but unsuccessful takedown only serves the company’s strategic interests.

As noted, McNeil says that in his experience, this kind of dithering by Instagram is relatively new behavior, which at least implies that a policy change may be lurking behind the opaque walls at Facebook HQ. But whether the scattershot responses he received are the result of executive suite decisions or cubicle level incompetence, what they do indicate at scale is that Instagram is guilty of noncompliance with the statutory conditions required to retain its “safe harbor” immunity under Section 512.

Assuming McNeil’s experience is not unique (why would it be?) and thousands of copyright owners are lately receiving emails alleging that they are not rightfully authorized to submit takedowns, this begins to look like a strategy which endeavors to avoid expeditious removal of material upon receipt of properly filed takedown requests. If that is what’s going on, Facebook’s attorneys should perhaps look at what happened to Cox Communications when it lost its §512 immunity and ask themselves, WTF are they doing? Because demanding that a notice filer prove, beyond the statutory requirements of the DMCA, that he is the copyright owner, does appear to be an even brighter red-line violation of the conditional safe harbor than Cox’s pattern of avoiding account termination in its litigations with record labels.

Still, McNeil did provide the Insta-minions with the most reasonable proof of ownership available (short of a full-blown discovery process), despite being under no obligation to do so. He sent them side-by-side composites of his Harryhausen photograph next to the relevant Instagram copies, and … nothing. Just a repeat of the same allegations that perhaps McNeil is not telling the truth about being the owner of the work.

The fact that this legally unfounded insult is aggravated by the inconsistency that twenty-two takedowns of the same image were eventually honored is enough to make a sane man weep into the pages of his Kafka anthology. So, when the anti-copyright crowd says that DMCA abuse is a problem, they’re not wrong, except of course about who they say is abusing it.  

Also see story at Fstoppers.


Photo by: pareap

Podcast – The Multi-Billion-Dollar Piracy Industry with Tom Galvin of Digital Citizens Alliance

In this episode, I speak with Tom Galvin, CEO of Digital Citizens Alliance, about piracy of creative works and DCA’s latest report, issued this month in collaboration with the research group White Bullet. The report, entitled Breaking Bad(s): How Advertiser-Supported Piracy Helps Fuel a Booming Multi-Billion Dollar Illegal Market, reveals that piracy is a highly profitable criminal enterprise and is intertwined with other forms of cyber-crime—from personal identity theft to national security

Piracy of creative works like motion pictures, TV shows, music, and live sports is a vast and growing criminal enterprise. In its latest report, Digital Citizens Alliance estimates the combined advertising and subscription revenue generated by piracy is at least $2.34 billion annually. Meanwhile, in addition to its ill-effects on the creators whose works are pirated and the online advertising ecosystem, piracy plays a key role in fostering other forms of cyber crime.

Episode Contents

  • 01:52 – Breaking Bad(s) Report Overview
  • 04:05 – Ad and subscription supported piracy
  • 06:49 – The online advertising ecosystem.
  • 08:49 – Some successful mitigation since 2014.
  • 11:14 – The downsides of piracy for brands.
  • 15:10 – Major brands found were Amazon, Facebook, & Google.
  • 18:01 – It is possible to do something.
  • 19:24 – Advertiser pressure to get ad tech to clean up its act.
  • 21:09 – Dangers to the consumer.
  • 27:13 – Why aren’t the hazards deterrents?
  • 30:30 – Drive-by malware.
  • 32:07 – Piracy is a vertical for broader criminal enterprise.
  • 33:26 – What about solutions.
  • 37:33 – Even if you don’t care about copyright owners…
  • 40:30 – Intersection with disinformation campaigns?