ISP fails to dismiss (c) and CMI claims based on watermarked photo

Goldstein v. Metropolitan Regional Information Systems, Inc., 2016 WL 4257457, No. TDC-15-2400 (D. Md. Aug. 11, 2016)

Goldstein is a professional photographer who registered a copyright in a 2007 photograph he took of the Silver Spring Metro Station, which appears on his website with the watermark “© www.goldsteinphoto.com” centered at the bottom of the image.  Defendant MRIS runs an online real estate listing service that allows subscribers, mainly real estate brokers, to post listings for available properties for a fee. Subscribers agree to assign to MRIS the copyright in any photograph they upload to the database.  In 2014, Goldstein learned that his Metro Photograph had been uploaded to the MRIS database, still with Goldstein’s original watermark, but also the additional watermark “© 2013 MRIS” or “© 2014 MRIS,” depending on the year the image was uploaded.  Despite notice from Goldstein’s attorney, the photo allegedly continued to be displayed on/uploaded to the MRIS site during 2014 and continuing into 2015, at which point the MRIS watermark was updated to “© 2015 MRIS.” Goldstein alleged that, as a result, the Metro Photograph was uploaded to additional real estate websites and used in additional real estate promotional materials without his permission.

Goldstein sued for infringement, contributory infringement, violation of the DMCA’s CMI provisions, and violation of the Lanham Act. Easiest bit first: Goldstein alleged that MRIS’s addition of its own copyright watermark to the Metro Photograph was a false representation under the Lanham Act. Although this sounds like a false advertising claim, the court found it Dastar-barred: Dastar held that “claims of false advertising and trademark infringement under the Lanham Act do not overlap with copyright claims” because “origin” means physical origin, and because otherwise plaintiffs could create a kind of perpetual copyright or patent.  “Thus, ‘creative’ works, those artistic creations that fall within the ambit of copyright law, necessarily fall outside the scope of the Lanham Act.”  This is overstated in some significant ways, but query whether anyone’s found a way to say it better.

Direct infringement: MRIS argued that Goldstein didn’t allege “volitional conduct” by MRIS, relying on Costar Group, Inc. v. LoopNet, Inc., 373 F.3d 544 (4th Cir. 2004). Direct infringement requires “actual infringing conduct with a nexus sufficiently close and causal to the illegal copying that one could conclude that the machine owner himself trespassed on the exclusive domain of the copyright owner.”  In Costar, LoopNet was a passive ISP that could not be held liable for direct infringement of Costar’s copyright when Costar’s unmarked photographs appeared on its website, even though it engaged in cursory screening of photos.  The court here, however, distinguished Costar as having been decided on summary judgment.  Goldstein alleged that MRIS or an agent of MRIS uploaded the Metro Photograph.   In one exhibit, the photo appears in a slide in a slideshow that was available on the MRIS website, which didn’t contain any reference to a particular real estate agent or broker, or to a particular property. “Drawing all inferences in Goldstein’s favor, this Exhibit, particularly with the markings added by MRIS, supports the reasonable inference that MRIS or its agent, rather than an outside user of MRIS, engaged in direct infringement by copying Goldstein’s copyrighted photograph to its website.”

So far, so good, but the court goes on to say more things that ought to be disturbing.  The court also distinguished Costar because “(1) when MRIS received it, the Metro Photograph, on its face, was marked as copyrighted by Goldstein; (2) MRIS, unlike LoopNet, took the affirmative step of marking the Metro Photograph as subject to its own copyright; and (3) MRIS then copyrighted its entire website, including the Metro Photograph.”  [The court here means “registered” for (3), and this conflation pervades the opinion, including the treatment of (1)—after 1978, all original works are “born” copyrighted, so if you see a modern photo you know it’s “copyrighted,” though you may not know who owns the copyright or whether it has been licensed or whether a fair use is being made.]   (2) is probably bad business practice, but it’s likely as automatic as many other activities that have been held not to strip ISPs of the DMCA safe harbor, when the court gets around to that.  Anyway, I’m not sure why any of these three facts makes MRIS’s conduct “volitional” if Loopnet’s wasn’t.  The court said that “MRIS doubly asserted a copyright in the Metro Photograph by stamping that facially copyrighted photograph with its own copyright markings and then copyrighting the entire website containing that photograph,” but that’s still not “volitional” conduct implicating the rights protected by copyright, under Costar.

Further comment: It seems really, really unlikely that MRIS uploaded the photos, as opposed to a MRIS user.  Should my prediction prove true, should MRIS get its attorneys’ fees for defending the direct infringement claim?

Contributory/vicarious infringement: contributory sufficiently pled; vicarious not.  MRIS knew or had reason to know of Goldstein’s copyright, because “[w]here works contain copyright notices within them, as here, it is difficult to argue that a defendant did not know that the works were copyrighted,” and when MRIS got Goldstein’s C&D, it knew or had reason to know of the infringing use.  And

MRIS’s practice of adding its own copyright to the Metro Photograph, a practice it continued even after being placed on notice that the Metro Photograph was being used on its website without Goldstein’s permission, can reasonably be considered ‘conduct that encourages or assists the infringement.’ For example, the addition of the MRIS copyright markings could lead MRIS users to believe that they had a license to use the Metro Photograph in their listings as part of their subscription to the MRIS service.

Though MRIS asserted that it promptly removed each reappearance of the photo, that wasn’t enough on a motion to dismiss, and its continued reappearance was enough for contributory infringement [in the absence of the DMCA].

Vicarious infringement/inducement: MRIS likely had the right and ability to control the infringing activity and charged a fee to post listings on its website, “the website has as its main purpose the posting of real estate listings, a purpose distinct from trafficking in infringing material. It is therefore not reasonable to conclude that the availability of infringing photographs such as the Metro Photograph drew customers to subscribe to the MRIS service.” The inducement claim failed for similar reasons.

Compliance with the DMCA, the court further held, was an affirmative defense not suitable for resolution on a motion to dismiss.  And here it gets even worse, though presumably the court could at least fix this on summary judgment: the court suggests that the copyright notice on the photo might trigger actual knowledge that using the photo would infringe “a copyright.”  How is that actual notice, especially given an automated upload?  Moreover, the statutory provision that noncompliant DMCA notices don’t give actual knowledge would seem to preclude this reasoning, since a watermark is a far cry from a DMCA notice with all the relevant information.  And again, the court suggested that MRIS’s addition of its own watermark went beyond “storage at the direction of a user of material that resides on a system or network” to “affirmative appropriation of another’s copyright.”

CMI violations: Goldstein alleged violations of § 1202(a) and (b). §1202(a)  makes it unlawful for a person to “knowingly and with the intent to induce, enable, facilitate or conceal infringement provide copyright management information that is false.” §1202(b) prohibits a person, “without the authority of the copyright owner or the law,” either to “intentionally remove or alter any copyright management information” or to “distribute” work “knowing that copyright management information has been removed or altered …knowing or ... having reasonable grounds to know that it will induce, enable, facilitate, or conceal an infringement.”  Again, the complaint adequately stated a claim, though the court commented that Goldstein could only recover once for a §1202 violation.

MRIS unconvincingly argued that the watermark wasn’t CMI because it didn’t meet the requirements for a full copyright “notice.” §1202 doesn’t require that.  §1202(a)’s requirements were satisfied because MRIS allegedly added its own copyright watermark to the photo without owning the copyright.  MRIS argued that it had a copyright in the database as a whole, thus had a good faith belief that its watermark wasn’t false.  That didn’t work because of (1) Goldstein’s watermark and (2) the procedural posture of a motion to dismiss.  MRIS argued that its watermark wasn’t added knowingly and with the intent to facilitate infringement, because it was added “automatically” to all images uploaded to their site, but again that’s not an argument that works on a motion to dismiss, and anyway Goldstein’s C&D provided actual knowledge.

§1202(b): MRIS argued that it didn’t “intentionally remove or alter any copyright management information,” since Goldstein’s watermark was left intact.  The court, however, declined to dismiss Goldstein’s claim that adding CMI was “constructive” alteration of his own CMI, since it had “more visual impact” and “undercut[]” the message that Goldstein owned the copyright. There was no case law either way on constructive alteration of CMI, but, “[p]articularly where MRIS’s copyright mark was placed immediately before Goldstein’s copyright mark and used more recent dates, that mark could be construed as trumping, diluting, or superseding, and thus altering, Goldstein’s CMI.”