“Rappers are skilled in poetry and rhythm—not necessarily in proper copyright registration procedures.” With that comment, the Court of Appeals for the Eleventh Circuit reverses a decision from the District Court for the Southern District of Florida that invalidated three copyright registrations for the same work and consequently dismissing the lawsuit. You can read my post on the district court case here, with the gory details on what went wrong with each registration.
Before digging into the appeals court decision, a little history lesson first. The Copyright Act was amended in 2008 to codify a standard for invalidating copyright registrations. The House Report at 24 explained the reason it was added:
It has also been argued in litigation that a mistake in the registration documents, such as checking the wrong box on the registration form, renders a registration invalid and thus forecloses the availability of statutory damages. To prevent intellectual property thieves from exploiting this potential loophole, the Act makes clear that a registration containing inaccuracies will satisfy the registration requirements of the Copyright Act unless the mistake was knowingly made and the inaccuracy, if known, would have caused the Register of Copyrights to refuse the registration. And in cases where mistakes in a copyright registration are alleged, courts will be required to seek the advice of the Register of Copyrights as to whether the asserted mistake, if known at the time of application, would have caused the Copyright Office to refuse registration.
These additions are now subparts (b) and (c) of section 411. The Copyright Office, in its 2008 Annual Report, p. 13, commented “it amends section 411 of the copyright law to codify the doctrine of fraud on the Copyright Office in the registration process.”
But if the intent was to make it more difficult to invalidate a registration, the provision, in my opinion, has utterly failed. Courts readily invalidate registrations, and this opinion may illuminate why.
With that, we reach the opinion. Section 411 says that one can invalidate a registration if “inaccurate information was included on the application for copyright registration with knowledge that it was inaccurate” and the inaccuracy would have caused the Register to refuse the registration. The copyright owners, William L. Roberts, who is rapper Rick Ross, Andrew Harr and Jermaine Jackson, argued in the district court without success that this is a fraud standard, with the district court responding “Title 17 U.S.C. § 411(b) says nothing about an intent to defraud the Copyright Office. Rather, the plain text of the statute requires only that the application be made ‘with knowledge that it was inaccurate.’”
The rappers were successful, though, in the appeals court. 11th Circuit precedent, both before and after 2008, had a scienter requirement, that “intentional or purposeful concealment of relevant information” is required for the invalidation of a registration. The court helpfully summarized the standard:
(1) the application must contain inaccuracies, (2) the inaccuracies must be material, and (3) the applicant must have the required scienter of intentional or purposeful concealment.
There was no intent to conceal here:
This is not a case where Rapper A attended a Rapper B concert, heard a delightful song, stole the composition, and fraudulently registered it with the Copyright Office—far from it. There is no dispute by any party that Appellants authored and created Hustlin’, and there is no dispute that they continue to receive the writers’ share of royalties from their musical composition. Furthermore, Appellees never proffered any argument or theory as to why Appellants would attempt to deceive the Copyright Office, when they are, in fact, the undisputed authors.
As indicated by the absence of any sort of motive for deception, the errors made in each of the registrations were done in good faith. As portions of the ownership interest were acquired by record companies, those companies—incorrectly, but in good faith—filed for a new registration to protect their newly acquired interests presumably under the assumption that no previous registration had been filed.
The failure of the first registration to correctly assert a published work on the basis of promotional phonorecords provided to disc jockeys—as opposed to an unpublished work that was still awaiting album publication—lacks any sort of deceptive intent, especially since there is nothing to indicate that the registration would not have been approved as a published work. Furthermore, nothing of substance could be gained by listing the incorrect creation date of 2006 instead of 2005 on the latter two registrations. Considering that the album publication occurred in 2006, it seems that an understandable—albeit incorrect—definition of publication persisted in the second and third registration.
The appeals court’s reasoning is impeccable, other than the statutory word choice “knowledge” is a far cry from “intentional concealment.” That is demonstrated by the appeals court’s somewhat glib re-writing of a Copyright Office regulation:
And while “[a]s a general rule only one copyright registration can be made for the same version of a particular work,” specific exceptions are recognized by a federal regulation. 37 C.F.R. § 202.3(b)(11). In addition to these three exceptions contained in a federal regulation, logic would dictate that Original Appalachian and St. Luke’s would support at least one more—a good faith, redundant registration for a published work.
I hope this is the correct legal standard. I do believe that the now-inevitable collateral attack on a copyright registration before even reaching infringment is wasteful, time-consuming and makes registration a trap for the unwary. There are some questions on the application for which the Copyright Office won’t even provide guidance, like whether a website is “published.” Expecting a copyright applicant to get it absolutely perfect on the law and the facts every time is unworkable. But I’m not too optimistic; the court clearly read an element, scienter, into the statute that just isn’t there.
There are also a couple of bonus issues in the opinion. One problem is that the Register’s opinion is incomplete. During the registration process, a copyright examiner may identify a problem and contact the applicant to get it corrected. But of course, by the time we get to an infringement proceeding the error is unrecoverable. There were several errors in the registrations in this case that may have been correctable had the applicants been given the opportunity:
The Copyright Office responded [in its § 411(c) response] that it “would have refused to issue a registration for an unpublished work” had it known of the publication/distribution information for the first registration. However, the Copyright Office’s response does not indicate whether it would have permitted the first registration as a published work.[fn4] Additionally, the Copyright Office advised that it “would have refused [the second] registration” had it known of the incorrect creation date. Lastly, the Copyright Office stated that it “would have refused [the third] registration” had it known of the incorrect creation date or the prior registration for a published work.…
[fn4] Absent from the Copyright Office’s response is any indication as to whether any of the inaccuracies from the three registrations would have been returned to the applicant for remediation, and whether that remediation would have caused the Copyright Office to subsequently accept the registrations during the application period. Although not dispositive of the issue of materiality, this context is relevant.
See U.S. Copyright Office, Compendium of Copyright Practices (3d ed. 2014) (“Compendium (Third)”).
The Copyright Office does not generally question facts alleged in an application “unless they are implausible or conflict with information provided elsewhere.” Compendium (Third) § 1904.2. While this explains the Copyright Office’s failure to identify the inaccuracies regarding publication and creation date, it does not explain why the Office failed to contemporaneously detect that the same song by the same authors had previously been granted registration(s).
The Copyright Office’s response itself notes that examiners often perform reviews and correspond with filers to correct inaccuracies such as those present in the Hustlin’ registrations, see Resp. of the Register of Copyrights, App. Doc. 383, at 4-5, including incorrect years of creation, see Compendium (Third) § 611.4, previous registrations, see Compendium (Third) § 621, and whether works have been published, see Compendium (Third) § 1904.3.
In instances of these same errors, the Copyright Office “provides the applicant an opportunity to correct the error or verify the facts.” Resp. of the Register of Copyrights, App. Doc. 383, at 5; see also Compendium (Third) § 605.6 (providing applicants 20 days for email responses or 45 days for postal mail responses).
(Emphasis in original.)
There was also an interesting procedural holding in this case. The appeals court relates that neither party contested the plaintiff’s ownership or validity of the registration until the court raised it sua sponte. The court of appeals held the district court acted improperly in doing so:
While the Supreme Court has declined to address whether “district courts may or should enforce [the registration precondition] sua sponte,” it did determine that “a copyright holder’s failure to comply with § 411(a)’s registration requirement [does not] deprive a federal court of jurisdiction to adjudicate his copyright infringement claim.” Logically, if failure to register does not eliminate subject matter jurisdiction, an improper registration would not either. Thus, the district court had jurisdiction.
In any event, the district court’s review of validity was clearly not an issue of subject matter jurisdiction, but rather the determination of an affirmative defense. Cf. Original Appalachian, 684 F.2d at 821, 827-28 (“While the burden of persuasion as to the validity of the copyright rests with the plaintiff in an infringement action, once he produces a copyright certificate he establishes a prima facie case of validity of his copyright and the burden of production shifts to the defendant to introduce evidence of invalidity.”) (internal citations omitted); see also Bateman v. Mnemonics, Inc., 79 F.3d 1532, 1541 (11th Cir. 1996) (“Once the plaintiff produces a certificate of copyright, the burden shifts to the defendant to demonstrate why the claim of copyright is invalid.”). Correspondingly, failure to plead an affirmative defense typically results in waiver of that defense. See Latimer v. Roaring Toyz, Inc., 601 F.3d 1224, 1239-40 (11th Cir. 2010). Courts “generally lack the ability to raise an affirmative defense sua sponte” with minor exceptions that are not relevant to copyright infringement actions. See id. at 1240 (finding that the district court erred by raising a “fair use” defense sua sponte in a copyright infringement action). Here, the district court—and not the defense—raised the issue of registration validity, and thus it erred in the manner of its review.
(Brackets in original.) It may be a stretch to say that invalidity of a registration is an affirmative defense (one of the cases refer to a shifting burden of production, not burden of persuasion, and it isn’t one of the many, many defenses listed in the Copyright Act), but it is for sure not jurisdictional, Reed Elsevier, Inc. v. Muchnick, 559 U.S. 154, 164-65 (2010), which the court is obliged to consider. But it’s an interesting thought.
A lot of meat in this opinion, all good from an IRL perspective. Here’s hoping it sticks.
Roberts v. Gordy, No. 16-12284 (11th Cir. Dec. 15, 2017).
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.