A panel of the Ninth Circuit has affirmed that the “server test” remains the law of the land - at least in that circuit, and at least for now.

Hunley v. Instagram

Photographers Alexis Hunley and Matthew Brauer (together, “Plaintiffs”) regularly post their work on Instagram. In 2016, Time embedded one of Brauer’s posts (featuring a photo of Hillary Clinton) in an article entitled “These Photographers Are Covering the Presidential Campaign on Instagram.” In 2020, Buzzfeed embedded a post published by Hunley (documenting a Black Lives Matter protest) in an article entitled “17 Powerful Pictures Of The Protests Through The Eyes of Black Photographers.” Buzzfeed and Time used Instagram’s embedding functionality and did not seek permission from the plaintiffs. (Images of the articles, as they originally appeared, are reproduced in the court’s opinion.)

As we have explained before (see here), a website developer can use HTML code (an embed code) to incorporate content (including photography) from another website rather than uploading a copy of that content directly to the server that hosts the developer’s website. When an individual visits a website that includes an embed code, the visitor’s internet browser is directed to retrieve the embedded content from the third-party server and display it on the website. As a result, although the visitor sees the embedded content on the visited website, that content actually is hosted on a third-party’s server and not on the server that hosts the visited website. Instagram makes it easy for developers to embed social posts by following these simple steps.

Plaintiffs sued Instagram (but not Buzzfeed or Time) in the Northern District of California on behalf of a class of similarly situated Instagram users, alleging that (1) when Buzzfeed and Time used Instagram's embedding functionality to display the photos on their websites, without the plaintiffs’ permission, they were guilty of direct copyright infringement of the plaintiffs’ exclusive right to publicly display the photos, and (2) Instagram was liable for secondary copyright infringement because it materially contributed to and induced that infringement by making its embedding tool available to Buzzfeed, Time, and the rest of the world. The plaintiffs conceded that Instagram was not liable for direct copyright infringement since each of them, via Instagram’s Terms of Service, had granted to Instagram a broad license to display their photos on Instagram.

District Judge Charles Breyer dismissed Plaintiffs’ claims twice, concluding each time that (1) because the plaintiffs’ photos were hosted on Instagram’s servers rather those of Buzzfeed and Time, any claim that Buzzfeed and Time were direct infringers was foreclosed by the server test, adopted by the Ninth Circuit in Perfect 10 v. Amazon, 508 F.3d 1146 (9th Cir. 2007), and (2) since Buzzfeed and Time were not direct infringers, Instagram could not be liable on secondary liability theories. (We covered the district court’s decisions here and here.) In its second opinion, the district court noted that since the plaintiffs clearly were of the view that "the Ninth Circuit's server test misinterprets copyright law," they were "free to present that argument to the Ninth Circuit." And that is precisely what the plaintiffs did.

Perfect 10 and the Server Test

In Perfect 10, adult website Perfect 10 (once famous for featuring models who had not undergone plastic surgery) sued Google alleging, among other things, that Google had directly infringed upon its copyrights (specifically, the public display right) when, as part of image search results, Google made available to its users embedded, full-sized images of Perfect 10’s copyrighted photographs. The Perfect 10 court held that, because Google had embedded these images from third-party servers, Google had not directly infringed upon Perfect 10’s display right. Parsing through the dense thicket of statutory definitions in the Copyright Act, the court concluded that a website, like Google, could not infringe upon a photographer’s public display right unless the website transmitted a ”copy” of the photograph to end users. Since Google did not store (or fix) a copy of Perfect 10’s images on Google’s own servers, it did not transmit a copy of those images to the public; instead, it transmitted only HTML code (i.e., embed code), which was not the same thing:

“Instead of communicating a copy of the image, Google provides HTML instructions that direct a user’s browser to a website publisher’s computer that stores the full-size photographic image. Providing these HTML instructions is not equivalent to showing a copy. First, the HTML instructions are lines of text, not a photographic image. Second, HTML instructions do not themselves cause infringing images to appear on the user’s computer screen. The HTML merely gives the address of the image to the user’s browser. The browser then interacts with the computer that stores the infringing image. It is this interaction that causes an infringing image to appear on the user’s computer screen. Google may facilitate the user’s access to infringing images. However, such assistance raises only contributory liability issues and does not constitute direct infringement of the copyright owner’s display rights.”

And thus the subsequently named “server test” was born. But it lives only in the 9th Circuit. No other circuit court has explicitly adopted this test – and a handful of district courts outside the 9th Circuit have expressly rejected it (discussed below). Judge Posner in the Seventh Circuit applied somewhat similar reasoning to the server test in overturning a preliminary injunction in Flava Works v. Gunter, but did not actually endorse the test. Which takes us back to the 9th Circuit’s revisiting the server test in Hunley.

Hunley Panel Applies Perfect 10

On appeal, Plaintiffs tried, but failed, to convince the panel that Perfect 10 did not control the outcome of the case.

First, the panel reaffirmed Perfect 10’s holding that a website operator does not directly infringe upon a photographer’s exclusive right of public display unless it transmits a display of a copy of the plaintiffs’ work that is fixed on the website operator’s server:

For those interested in how the court in Perfect 10 (and the court in Hunley) parsed through the various statutory definitions in 17 U.S.C. § 101, this color-coded graphic may (or may not) be helpful:

Second, the panel rejected the plaintiffs’ argument that the server test should apply only “to search engines or other automated, algorithmic indexing platforms” and not to commercial websites (like Buzzfeed and Time) that embed specific images from specific social media accounts to illustrate articles without a license from the copyright owners. Because Perfect 10 was based on the “plain language” of the Copyright Act and Ninth Circuit precedent defining when a copy is “fixed” in a tangible medium of expression, the panel saw no basis for limiting Perfect 10 to any specific type of website. Plain language is plain language. As the court explained:

“We have not limited Perfect 10 to search engines, and it is too late to argue that it is so limited. The application of the Server Test depends on the method used for displaying a photo – not the context in which the photo is displayed. And the process used by BuzzFeed and Time to show users third-party Instagram content is the same process used by Google to show users third-party images: embedding. Nothing in Perfect 10 or the cases following it limits its application to search engines.”

Third, the panel acknowledged that a handful of courts (mostly district courts) in other circuits have rejected or limited the server test. See, e.g., McGucken v. Newsweek; Nicklen v. Sinclair Broad. Grp., Inc.; and Goldman v. Breitbart. (See also this post discussing these cases.) Needless to say, the Ninth Circuit was not bound to follow (and, perhaps, was not persuaded by the reasoning of) these out-of-circuit district judges.

Fourth, the plaintiffs contended that Perfect 10 misconstrued the Copyright Act in several ways. Among other things, the plaintiffs argued that Perfect 10’s holding that the right to publicly display is only infringed when a copy is made (i.e., when the right to reproduction is also infringed) conflated these two rights and rendered portions of the Copyright Act superfluous or insignificant. Plaintiffs also argued that the server test cannot be reconciled with other provisions of the Copyright Act that prohibit transmissions by a party, whether or not the party possesses or controls a copy of the work allegedly infringed. The panel held that these arguments were foreclosed by Perfect 10, even if they were not expressly addressed in that decision. “Even if we thought, in retrospect, that Perfect 10 created some inconsistencies with other provisions of the Copyright Act, we are not free to overrule Perfect 10 outside of an en banc proceeding unless there has been a change in the statute or an intervening Supreme Court decision.” Too little, too late.

Fifth, the panel rejected the plaintiffs’ argument that Perfect 10 was “effectively overruled” by SCOTUS’s decision in American Broadcasting Company v. Aereo, 573 U.S. 431 (2014). Aereo offered subscribers an ability to stream on their computers copyrighted programming that was broadcast over the air by television stations. A subscriber could visit Aereo’s website to select programming it wanted to view, and Aereo would tune an antenna dedicated solely to that subscriber to capture and record the programming (which was stored in a subscriber specific folder), which it would stream to the subscriber over the internet. Aereo established this individualized antenna framework to attempt to circumvent the broadcasters’ exclusive right to publicly perform the work; in other words, Aereo argued that any performance was not “public.” The broadcasters countered, however, that Aereo’s technological gymnastics amounted to form over substance, and Aereo’s actions nevertheless infringed the broadcasters’ public performance right. SCOTUS agreed, finding that Aereo was not merely providing hardware for a subscriber to record (time shift) and then watch on their computers (space shift); it also publicly performed the programming since it “transmit[ed] … a performance” of the programming to multiple subscribers, albeit through separate antennas.

Plaintiffs argued that the display right (at issue in this case and Perfect 10) and the performance right (at issue in Aereo) were “easily interchanged,” and that it defied logic to conclude that Aereo publicly performed the programming delivered to its subscribers, but Time and Buzzfeed did not publicly display photos embedded on their websites. As Plaintiffs put it: “it is the practical, functional perspective of the public viewer, and not hyper technicalities, that must determine whether a particular mode of content or delivery system is infringing or not.”

The Hunley panel concluded that Aereo did not overrule Perfect 10. “Perfect 10 and Aereo deal with separate provisions of the Copyright Act – Perfect 10 addressed the public display right, and Aereo concerned the public performance right.” This matters, since there are “significant” differences between the Copyright Act’s definitions of “display” and “perform”:

“Infringing the public display right requires an underlying copy. By definition, displaying a work publicly requires that the infringer display a copy of the work, id. § 101 (definition of “display”); and transmission of a display means that someone has transmitted a copy of the work “to the public.” Id. (definition of “To perform or display a work ‘publicly’”). However, to infringe the public performance right, the infringer need not show or perform a copy of the underlying work. Id. § 101 (definition of “perform”).”

Sixth, the panel suggested that the plaintiffs’ claim must fail because Time and Buzzfeed has not engaged in sufficient volitional conduct to be liable as a direct infringer:

“Our volitional conduct requirement draws a distinction between direct and secondary infringement that would likely foreclose direct liability for third-party embedders. And without direct infringement, Hunley’s secondary liability theories all fail.” (Emphasis added.)

(Observation:  It is one thing to say that Google’s conduct in Perfect 10 – which consisted of providing automated responses to users’ search queries that enabled users to locate images posted on third party websites – didn’t cross the threshold of volitional conduct to be actionable. However, reasonable minds can differ as to whether Buzzfeed and Time engaged in sufficient volitional conduct since they chose the specific images from the Plaintiffs’ social accounts and took all actions required to embed those images in their articles.)

Seventh, the panel rejected Plaintiffs’ argument that “user perception” provided a sufficient basis to find an embedder to be liable for direct infringement. Plaintiffs pointed out that the typical user would have no idea whether an image is hosted on the server of the website the user visited or a third-party server. As Plaintiffs’ attorneys put it, “[w]hat happens behind the curtain is . . . irrelevant to the consuming public, and so too should it be irrelevant in the eyes of the law.” This didn’t fly with the panel: “because the public display right requires that an infringer ‘display a copy,’ user perception is relevant to whether a copy is ‘fixed,’ but not sufficient to establish direct infringement absent the underlying display of a fixed copy.”

Finally, Plaintiffs, Instagram and various amici advanced many policy arguments in favor or against the server test. While these were “serious and well-argued,” the court was not the right forum to address those issues:

“If Hunley disagrees with our legal interpretation – either because our reading of Perfect 10 is wrong or because Perfect 10 itself was wrongly decided – Hunley can petition for en banc review to correct our mistakes. But we have no right “to judge the validity of those [] claims or to foresee the path of future technological development.” Aereo, 573 U.S. at 463 (Scalia, J., dissenting). Most obviously, Hunley can seek further review in the Supreme Court or legislative clarification in Congress.”

So … the server test holds for now in the Ninth Circuit alone. Surely this will not be the last word on the matter since the court invited Plaintiffs to seek en banc and Supreme Court review.

 Hunley v. Instagram, LLC, __ F. 4th __, No. 22-15293 (9th Cir. July 17, 2023)