©2015. Published in Landslide, Vol. 7, No. 3, January/February 2015, by the American Bar Association. Reproduced with permission. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association or the copyright holder.
To some, § 512 of the 1998 Digital Millennium Copyright Act (DMCA)1 is a lot like the South American cane toad. In the early part of the twentieth century, beetle populations were destroying sugar cane crops in parts of the United States and Australia. So, to solve that problem, thousands of cane toads were released into the wild: nearly foot-long ravenous omnivores with no natural predators, capable of exuding a venomous toxin sufficient to kill a large dog within 15 minutes. What could possibly be the downside?2 The sugar cane looks great, but where did the dog go?
Thus the analogy to the DMCA, which some argue is a copyright enforcement mechanism that has created nasty problems even while solving others. The DMCA does undeniably remedy some—even a lot of—copyright infringement, but it is often misused and abused. Is the specific statutory mechanism designed to deter misuse of the DMCA, i.e., § 512(f), up to the job?
The DMCA and Its Abuse
Section 512 of the DMCA essentially represents a deal between the government and the Internet industry. In return for immunity against most secondary copyright liability claims, § 512 mandates a notice and takedown regime to be paid for and administered by Internet service providers, with very little government involvement.
So, if someone online has pirated your copyrighted work—say a photograph—you can notify the alleged pirate’s Internet service provider and get the infringing copy taken down or the link to it disabled. Per § 512(c), this notification should identify your copyrighted photograph, provide the location of the infringement, and include a statement that you have “a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law.” If the alleged pirate disagrees, he or she can file a counternotice pursuant to § 512(g), including a sworn statement that he or she has a “good faith belief” that the takedown was “a result of mistake or misidentification.” At that point, provided you haven’t already filed suit, the photograph can go back up, and if you still want to fight about it you can always go to court.
Two things have made this much more complicated than it sounds. First: volume. Google alone received 230 million takedown notices in 2013 (and that volume appears to be growing at a scary rate).3 Second: abuse. Critics charge that § 512’s easy availability has allowed it to become a tool for the greedy to monopolize online speech, for the craven to censor political enemies, and for the ignorant to enforce nonexistent rights. When you couple this abuse with the aforementioned volume, the result is that even the most well-meaning Internet service providers are regularly taking down stuff that nobody ought to have asked them to take down in the first place, and that they might not have taken down had they had a chance to breathe. And usually, unless a counternotice is filed, it stays down.
It’s hard to pin down exactly what percentage of takedown notices qualify as abuse, but anecdotes abound. A recent amicus brief4 filed in the Ninth Circuit by Google, WordPress, and others lists a familiar parade of DMCA misuse horribles:
- A children’s book publisher sent notices to take down critical reviews about its books;
- A state governor submitted notices to remove his image from a political parody site;
- A model in a contract dispute with a photographer sought to remove images that had been taken and were owned by the photographer;
- A company in India falsely claimed to own a scientific article in order to have it removed from a WordPress blog;
- An actor sought removal of a Google Earth photograph of his residence, falsely claiming to be the rights holder;
- A company using forged customer testimonials on its website submitted a notice to take down an online article exposing the scam;
- Major television networks sent takedown notices targeting presidential campaign videos that included brief excerpts of news footage;
- A corporation submitted DMCA takedown notices to remove company documents posted by a whistleblower;
And my favorite . . .
- A frequent submitter of takedown notices sought to use the DMCA to erase an online discussion criticizing him for submitting overreaching DMCA notices.
Section 512(f) to the Rescue?
So, what’s the recourse? If you receive a bogus takedown notice, you can issue a counternotice and have the material put back up. But that doesn’t prevent or deter your malefactor from issuing another one every day until you just give up.
That’s where § 512(f) comes in, or was supposed to come in, depending on who you ask. Section 512(f) provides that, in the context of a takedown notice, “[a]ny person who knowingly materially misrepresents . . . that material or activity is infringing” is liable for damages, costs, and attorneys’ fees caused by the misrepresentation. One court called it a “reverse DMCA case.”5 According to a 1998 Senate report prior to passage of the DMCA, this dual mechanism—an easy takedown process coupled with the threat of punishment for abuse of that process—was supposed to provide a rapid response to potential infringement while at the same time deterring allegations that are knowingly false.6 It seems rather straightforward . . . until you try to define “knowingly.”
The Knowingly Unknown
The case with the biggest impact on the interpretation of “knowingly” was Rossi v. Motion Picture Ass’n of America.7 Michael J. Rossi operated the website internetmovies.com, which featured images from various movies, next to which it claimed to have “Full Length Downloadable Movies” and invited users to “Join to download full length movies online now!”8 A Motion Picture Association of America (MPAA) employee, noticing that among the films being advertised were those to which MPAA controlled the copyright, issued a takedown notice claiming that Rossi was distributing unauthorized copies of MPAA property.
However, the site actually did not offer full-length movies for download, or at least not the MPAA movies it claimed to offer. Had the MPAA employee gone through the trouble of joining Rossi’s service and trying to download the videos, the MPAA would have realized that there was no infringement. Rossi brought suit under § 512(f). He claimed that because takedown notices had to be issued in “good faith,” the MPAA was required to do some sort of reasonable investigation prior to issuing the notice, which it admittedly did not.
However, the Ninth Circuit disagreed, holding that “good faith” was a subjective standard, requiring no reasonable investigation. Therefore, the MPAA’s mistake was not “knowing.” Under § 512(f), a “copyright owner cannot be liable simply because an unknowing mistake is made, even if the copyright owner acted unreasonably in making the mistake.”9 “Rather,” the court concluded, “there must be a demonstration of some actual knowledge of misrepresentation on the part of the copyright owner.”10
So “good faith” is subjective—we can live with that; after all, the Ninth Circuit looked it up in Black’s Law Dictionary and it seemed to check out.11 But since when was the term “knowingly” completely subjective or limited to actual knowledge? Just a few months before the Ninth Circuit decided Rossi, a California federal court, also citing Black’s Law Dictionary, interpreted “knowingly” in § 512(f) to mean actual and constructive knowledge: that “a party actually knew, should have known if it acted with reasonable care or diligence, or would have had no substantial doubt had it been acting good faith.”12 And this was no outlier. In fact, only a couple years before Rossi, the Ninth Circuit had interpreted “knowingly” in pretty much the same way in the context of the False Claims Act, to include actual knowledge, deliberate ignorance, and reckless disregard.13
Why did the Ninth Circuit apply a different definition for § 512(f)? Some think that the Rossi court’s interpretation was caused by a mistaken transposition of the “good faith” standard in § 512(c) into the “knowingly” standard in § 512(f), resulting in the evisceration of an important safeguard against abuse. But others felt the holding was necessary and advisable. Under this view, the § 512(g) counternotice procedure is already remedy enough in most instances, and § 512(f) litigation should be an extremely rare bird, limited to cases of truly egregious bad acts—not mistakes. Whichever view is correct, Rossi undeniably declawed § 512(f), and it is still the most influential interpretation of the statute.
This One’s for You, Dr. Amy
In Rossi, the MPAA’s mistake was one of fact, i.e., the relatively clean question of whether the copyrighted work was being distributed at all. But what happens when the mistake is one of law? Say, for example, you don’t think something is fair use but you are wrong. Or you just don’t care whether it is fair use before issuing a takedown notice. Or maybe you don’t even know what fair use is.
In Tuteur v. Crossley-Corcoran, two bloggers—“Dr. Amy” and “the Feminist Breeder”—got into a nasty online flame war over the relative merits of giving birth at home.14 The Feminist Breeder escalated the argument by uploading a picture of herself giving “the finger” with the caption: “This one’s for you, Dr. Amy,” and further commenting that the photograph was “something else [for Dr. Amy] to go back to her blog and obsess about.”15 Dr. Amy responded in turn by republishing the Feminist Breeder’s photo on her own blog, along with additional commentary. Upon seeing her photo republished, the Feminist Breeder issued a DMCA takedown notice resulting in the disabling of Dr. Amy’s blog.
Dr. Amy filed suit against the Feminist Breeder for violation of § 512(f), and the Feminist Breeder moved to dismiss. Dr. Amy’s argument was that, although the copyright in the photograph clearly belonged to the Feminist Breeder, it was just as clear to an objectively reasonable person that Dr. Amy’s republishing of it was fair use. Therefore, the Feminist Breeder could not possibly have believed she had a viable copyright claim unless she utterly failed to even consider fair use before issuing the takedown notice.
The court agreed that Dr. Amy’s fair use claim was “plausible, and even dispositive.”16 However, relying on Rossi’s subjective standard, the court held that the fact that a takedown notice was objectively unreasonable was not sufficient evidence of a knowing misrepresentation. The statute did not require that copyright owners stop and consider defenses before issuing takedown notices, because that would defeat the “expeditious” nature of the remedy. Rather, what was required was a pleading indicating some evidence that the Feminist Breeder actually knew her claim was rotten.17
In one sense, this is a fair result. The Feminist Breeder is a lot of things: blogger, doula, activist, even the former bass player for the band Veruca Salt. But she is not a lawyer. So perhaps it would have been too much to make her conduct a fair use analysis prior to issuing a takedown notice. But Rossi dictated the same result in Dudnikov v. MGA Entertainment, Inc., where the court refused to hold an intellectual property lawyer to a higher standard after he had issued an incorrect takedown notice.18 The upshot is that it is exceedingly difficult to show that an individual acted “knowingly.” Does the same hold true for a corporation?
Let’s Not Go Crazy
In 2007, Stephanie Lenz uploaded to YouTube a short video of her toddler dancing. Playing on a radio in the background was a poor-quality 20-second audio clip of Prince’s Let’s Go Crazy. Universal Music Corp. issued a takedown notice, and the video was temporarily disabled.
After submitting a counternotice to YouTube and having the video restored, Lenz (with the help of the Electronic Frontier Foundation) brought suit against Universal under § 512(f). Lenz alleged that Universal knew her video was “self-evident” fair use, or at least failed to consider whether it was fair use, and only issued the takedown notice in deference to Prince’s outsized view of the scope of his intellectual property rights. In Lenz v. Universal Music Corp., the Northern District of California denied Universal’s motion to dismiss and held, even after Rossi, that in order to further the “good faith” provisions of the DMCA, a copyright owner had to consider the fair use doctrine before issuing a takedown notice.19
However, at some point between the pleadings and summary judgment, the Rossi holding sunk in. In 2013, the court retreated from its earlier position and held, “In light of Rossi, it appears that Universal’s mere failure to consider fair use would be insufficient to give rise to liability under § 512(f). Lenz thus must demonstrate that Universal had some actual knowledge that its Takedown Notice contained a material misrepresentation.”20
But this wasn’t a total victory for Universal. By this time, the court and Lenz were ready to play by Rossi’s rules. Lenz was able to develop evidence in discovery that Universal’s system for identifying online infringement was administered by employees with no legal training, using criteria that could not possibly identify fair use. Thus, Lenz alleged that Universal’s procedures for issuing takedown notices were so deficient that that it willfully blinded itself to the issue altogether, and this willful blindness was tantamount to actual knowledge in copyright law. Thus, Universal’s motion for summary judgment was denied. As of this writing, Lenz is on appeal to the Ninth Circuit on this very issue.
How to Survive in a Rossi Universe
Even though Rossi rendered § 512(f) less of a deterrent than it might have been, that doesn’t mean the statute is completely toothless. Here are a few situations that seem to be consistent with, or at least coexist with, the Rossi standard:
Despite the District of Massachusetts’s adoption of the Rossi standard in Tuteur v. Crossley-Corcoran, Dr. Amy was actually able to meet that standard in a manner sufficient to survive a motion to dismiss, thanks to some alleged inopportune admissions by the Feminist Breeder and her attorney. Dr. Amy was able to allege that (1) the Feminist Breeder’s attorney had at one point admitted to Dr. Amy’s attorney over the phone that she had no credible copyright claim; and (2) the Feminist Breeder implied on her blog that what she really wanted was not to protect a copyright but to force Dr. Amy “to agree to stop personally attacking me.” Thus, the court found that a “knowing and material misrepresentation” was adequately pleaded, and the case settled shortly thereafter.
In Rosen v. Hosting Services, Inc., a photographer issued a takedown notice, alleging that four of his photographs of model Daisy Fuentes were available on a website without his authorization.21 However, the photographer goofed—they weren’t pictures of Daisy Fuentes, but of actress Amy Weber. They were still his photographs, but nevertheless the court found that the incorrect descriptions could be found to be a knowing material misrepresentation under § 512(f), thus precluding summary judgment.
Use of Automated Systems
An area that is still very much developing is the intersection between § 512(f) and automated systems. Say you set up a computer to identify unauthorized uses of your copyrighted work online and automatically spit out takedown notices. How is a computer supposed to form a “good faith” belief that something is infringing under § 512(c)? On the other hand, how can a computer make a “knowingly” false statement under § 512(f)? In Disney Enterprises, Inc. v. Hotfile Corp., Warner Brothers had such a system, allegedly resulting in 400,000 takedown notices sent to Hotfile, about 600–900 of which were erroneous.22 The court pointed out that Warner’s system was set up to identify what the company owned, but not necessarily whether the use was infringing. Moreover, Warner’s reliance on automation appeared to skirt the obligation to form a good faith belief that infringement was occurring. That, combined with the apparently overzealous scope of the material identified, was sufficient for Hotfile’s § 512(f) claim to survive summary judgment.
Trademarks and the “Facebook Said I Could” Defense
If someone is using your trademark online without authorization, it may be tempting to utilize the simple and powerful procedures of the DMCA to deal with it. Don’t. Take Smith v. Summit Entertainment LLC, in which a songwriter decided to rebrand his song online as “inspired by the Twilight Saga.”23 This was a trademark matter—Summit didn’t claim to own a copyright to the song. But the Internet service provider only offered a copyright takedown option—nothing for trademarks—so Summit used the DMCA system to get the offending material taken down. The court held that the songwriter had sufficiently pleaded a “knowing misrepresentation” under § 512(f).
Similarly, in CrossFit, Inc. v. Alvies, CrossFit used Facebook’s DMCA reporting system to take down the Facebook page of an exercise guru who was using the CrossFit trademark without permission.24 CrossFit claimed that Facebook allowed both DMCA claims and trademark infringement claims to be reported through the same system, so it should not be liable under § 512(f). However, the court wasn’t buying (what I call) the “Facebook said I could” defense, and refused to dismiss the claim.
Inability to Articulate Entitlement to Copyright Protection
In Design Furnishings, Inc. v. Zen Path LLC, a furniture company issued takedown notices against its rival, which was apparently selling identical wicker furniture.25 In part because most furniture is denied copyright protection under the useful article doctrine, the rival asked the furniture company by e-mail for the basis of its claim to copyright protection. The furniture company never answered, and its lawyer was later unable to articulate for the court why the company thought the furniture was entitled to protection. The court held that this was sufficient indication of a knowing misrepresentation to justify a preliminary injunction against the furniture company’s issuance of further takedown notices.
The DMCA should be used cautiously in any dispute between potential co-owners of a work. In TD Bank, N.A. v. Hill, a bank executive authored a book about banking with the permission of his employer, pursuant to a series of contracts.26 When he left the bank and authored another similar book, the bank issued DMCA takedown notices to prevent its sale, claiming it infringed on the first book. The executive brought a claim under § 512(f), alleging that the contracts had made clear that he was the copyright owner of the initial book and the bank’s knowledge of those contracts rendered the takedown notice a “knowing misrepresentation.” The court agreed that the executive had properly stated a claim under § 512(f).
Finally, it is worth noting that § 512(f) also applies to knowing misrepresentations in counternotices. For example, in Shropshire v. Canning, songwriter “Dr. Elmo,” owner of the copyright to “Grandma Got Run Over by a Reindeer,” e-mailed YouTube user Aubrey Canning and asked her to remove a video containing his song.27 Canning acknowledged that it was Dr. Elmo’s song, but was otherwise defiant. Dr. Elmo issued a takedown notice, and Canning submitted a counternotice resulting in the allegedly infringing material staying online. Canning’s initial acknowledgment of Dr. Elmo’s rights appeared to conflict with her later counternotice statement that she had a “good faith belief” the takedown was the result of a “mistake or misidentification.” Thus, a knowing misrepresentation in the counternotice was adequately pleaded, and Dr. Elmo could proceed with his § 512(f) claim.
1. 17 U.S.C. § 512.
2. See Cane Toad, U.S. Dep’t of Agric. Nat’l Agric. Libr. http://www.invasivespeciesinfo.gov/aquatics/canetoad.shtml (last modified July 8, 2014); The Cane Toad (Bufo marinus), Australian Gov’t Dep’t of the Env’t (2010), http://www.environment.gov.au/biodiversity/invasive-species/publications/factsheet-cane-toad-bufo-marinus.
3. See Section 512 of Title 17: Hearing Before the Subcomm. on Courts, Intellectual Prop., and the Internet of the H. Comm. on the Judiciary, 113th Cong. 106 (2014) (statement of Katherine Oyama, Sr. Copyright Policy Counsel, Google Inc.), available at http://www.judiciary.house.gov/_cache/files/22c3acda-551c-41ba-b330-8dd251dd15fd/113-86-87151.pdf.
4. See Brief of Amici Curiae Automattic Inc.; Google Inc.; Twitter Inc.; and Tumblr, Inc. Supporting Plaintiff-Appellee-Cross-Appellant Lenz, Lenz v. Universal Music Corp., No. 13-16106 (9th Cir. Dec. 13, 2013), available at https://www.eff.org/files/2013/12/17/048_automattic_google_twitter_tumblr_amicus_brief_12.13.13.pdf.
5. Project DOD, Inc. v. Federici, No. 09-213-P-H, 2009 U.S. Dist. LEXIS 117401, at *1 (D. Me. Dec. 13, 2009).
6. S. Rep. No. 105-190, at 21, 49 (1998).
7. 391 F.3d 1000 (9th Cir. 2004).
8. Id. at 1001–02.
9. Id. at 1005.
10. Id. (emphasis added).
11. Id. at 1004 n.5 (citing Black’s Law Dictionary 701 (7th ed. 1999)).
12. Online Policy Grp. v. Diebold, Inc., 337 F. Supp. 2d 1195, 1204 (N.D. Cal. 2004) (citing Black’s Law Dictionary (8th ed. 2004)).
13. United States ex rel. Rakow v. Pro Builders Corp., 37 F. App’x 930, 930 (9th Cir. 2002); see also United States v. Shannon, 137 F.3d 1112, 1117 (9th Cir. 1998) (interpreting federal criminal law); Cal. Dairies, Inc. v. RSUI Indem. Co., 617 F. Supp. 2d 1023, 1047 (E.D. Cal. 2009) (interpreting California law).
14. Tuteur II, 961 F. Supp. 2d 333 (D. Mass. 2013).
15. Id. at 335.
16. Tuteur v. Crosley-Corcoran (Tuteur I), 961 F. Supp. 2d 329, 332 (D. Mass. 2013).
17. Tuteur II, 961 F. Supp. 2d at 343–44.
18. 410 F. Supp. 2d 1010 (D. Colo. 2005).
19. 572 F. Supp. 2d 1150 (N.D. Cal. 2008).
20. Lenz v. Universal Music Corp., No. 5:07-cv-03783-JF, 2013 U.S. Dist. LEXIS 9799, at *18–19 (N.D. Cal. Jan. 24, 2013).
21. 771 F. Supp. 2d 1219 (C.D. Cal. 2010).
22. No. 11-20427-CIV, 2013 U.S. Dist. LEXIS 172339 (S.D. Fla. Sept. 20, 2013).
23. No. 3:11CV348, 2011 U.S. Dist. LEXIS 60246 (N.D. Ohio June 6, 2011).
24. 109 U.S.P.Q.2d 2049 (N.D. Cal. 2014).
25. 97 U.S.P.Q.2d 1284 (E.D. Cal. 2010).
26. No. 12-7188, 2014 U.S. Dist. LEXIS 13519 (D.N.J. Feb. 3, 2014).
27. 809 F. Supp. 2d 1139 (N.D. Cal. 2011).