Ninth Circuit Reaffirms the “Server Test” for Direct Infringement of the Public Display Right — Hunley v. Instagram, LLC (Guest Blog Post)

By Guest Blogger Tyler Ochoa

Recently, the Ninth Circuit reaffirmed what has become known as the “server test”: in order to be held directly liable for violating the public display right, the alleged infringer must have a fixed “copy” of the work stored on a server in its possession or control.  Hunley v. Instagram, LLC, 2023 WL 4554649 (9th Cir. July 17, 2023).

Hunley deepens a split between the Ninth Circuit and district courts in other circuits (primarily the Second Circuit) which have rejected the application of the “server” test to so-called “embedded” images: HTML code that directs the user’s browser to display the linked-to image without further action from the user, rather than merely displaying a “hyperlink” that the user must click on in order to display the linked image. The Ninth Circuit panel also strongly hinted that if the plaintiff wanted to pursue its arguments, it should file a petition for rehearing en banc. For those reasons, I strongly suspect this will not be the last we hear of this case.

The Facts

The facts are relatively straightforward. The plaintiffs (Alexis Hunley and Matthew Scott Brauer) are photographers who posted some of their own photos to Instagram.

In 2016, Time “embedded” one of Brauer’s Instagram posts, featuring one of his photos of Hilary Clinton, in its entirety (preserving his username or “handle”, his caption, and his links and hashtags).

In 2020, BuzzFeed “embedded” one of Hunley’s Instagram posts, featuring one of her photos of a Black Lives Matter protest, in its entirety (again, preserving her username, caption, and hashtags). (See the screenshot on the right). Neither Time nor BuzzFeed sought or obtained authorization from the photographers to do so.

Neither Time nor BuzzFeed was named as a defendant. Instead, Hunley and Brauer filed a class-action lawsuit against Instagram, alleging that Instagram was vicariously liable for, or was liable for encouraging or contributing to, the alleged direct infringement by others, by providing an “embedding” tool that easily could be used to facilitate public display of their photos.

(One is tempted to characterize these plaintiffs as copyright “trolls.” It is certainly within the definition of chutzpah to publicly display your own work on the Internet, visible for anyone to see for free and without further conditions, and then to complain that others are helping people find your work by linking to it. On the other hand, professional photographers have to make a living by licensing their photos to publishers (and other intermediaries) who are willing to pay them to reproduce and display those photos. How are photographers supposed to get the attention of those publishers without displaying samples of their work? In other words, the gist of the case is whether the photographers surrender their right to exclude others by voluntarily posting their own photos to social media. As a practical matter, the answer is certainly yes; an open system is built into the design of the internet. The question is whether the same is true as a legal matter.)

It is hornbook law that secondary liability of all kinds requires proof that direct infringement is occurring. But Time and BuzzFeed did not directly “reproduce” the posts by embedding them; they merely used HTML code that referenced the online locations of the Instagram posts that contained those photos. That HTML code directed the user’s browser software to retrieve the 1s and 0s stored at that online location and to display the results to the user. The legal question is whether third parties (such as Time and BuzzFeed) “display a copy of the copyrighted work publicly” [17 U.S.C. §106(5)] by deliberately “embedding” such HTML code in their own posts.

Legal Background: The Public Display Right

Generally speaking, the U.S. Copyright Act grants authors five exclusive rights: “to reproduce the copyrighted work in copies or phonorecords”, “to prepare derivative works based on the copyrighted work,” “to distribute copies or phonorecords of the copyrighted work to the public,” “to perform the copy­righted work publicly,” and “to display the copyrighted work publicly.” 17 U.S.C. §106. This case concerns only the public display right.

“To ‘display’ a work means to show a copy of it, either directly or by means of a film, slide, television image, or any other device or process.”  17 U.S.C. §101. A work can be displayed “publicly” in one of four ways; the relevant one here is “to transmit or otherwise communicate a … display of the work … to the public, by means of any device or process, whether the members of the public capable of receiving the performance or display receive it in the same place or in separate places and at the same time or at different times.” Id.

A “copy” is defined as “[a] material object[] … in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.” Id. Copyright thus distinguishes between the “work” (here, an intangible image and accompanying text) and fixed “copies” of a work (such as a negative, or a print). A work is “fixed” “when its embodiment in a copy … is sufficiently permanent or stable to permit it to be perceived, reproduced, or otherwise communicated for a period of more than transitory duration.”  The 1s and 0s that are magnetically encoded on a hard disk or other electronic medium are a digital representation of the work that is “fixed” and that “can be perceived, reproduced, or otherwise communicated … with the aid of a machine or device.” Id.  Thus, an electronic file (or, more accurately, a particular segment of the hard drive) is a “copy” within the meaning of the statute.

Combining these definitions makes it clear that a public display occurs when an Instagram post is embedded in another webpage. The electronic file that resides on Instagram’s servers is a “copy” of the work. That copy is “displayed” to others: it is “show[n] … by means of … [a] device or process,” namely, the embedding of HTML code and subsequent transmission. Finally, it is displayed “publicly”: the 1s and 0s comprising the electronic file are “transmit or otherwise communicate[d] … to the public,” even though members of the public “receive it … in separate places” (their respective homes or offices) and “at different times” (asynchronously).

The next question is: who is legally responsible for this public display?

Legal Background: Direct and Secondary Liability

The Copyright Act gives the author the exclusive right “to do and to authorize” any of the five exclusive rights. 17 U.S.C. §106. The legislative history indicates that “Use of the phrase ‘to authorize’ is intended to avoid any questions as to the liability of contributory infringers. For example, a person who lawfully acquires an authorized copy of a motion picture would be an infringer if he or she engages in the business of renting it to others for purposes of unauthorized public performances.” H.R. Rep. 94-1476, at 61 (1976). The legislative history also expressed an intent to retain existing case law on vicarious liability of a principal for the acts of its agents, including independent contractors.  H.R. Rep. 94-1476, at 159-60.

As a result, courts have recognized three types of liability for copyright infringement: direct infringement (“to do” any one of the five exclusive rights), contributory infringement (“to authorize” another to directly infringe) and vicarious liability (based on the relationship between the direct infringer and the person sought to be held liable). While “direct” liability traditionally is regarded as a strict liability tort, the two types of secondary liability require additional elements. “[O]ne who, with knowledge of the infringing activity, induces, causes or materially contributes to the infringing conduct of another, may be held liable as a ‘contributory’ infringer.” Gershwin Pub. Corp. v. Columbia Artists Management, Inc., 443 F. 2d 1159, 1162 (2d Cir. 1971). And “even in the absence of an employer-employee relationship[,] one may be vicariously liable if [one] has the right and ability to supervise the infringing activity and also has a direct financial interest in such activities.” Id. These definitions have also been adopted by the Ninth Circuit, see Fonovisa, Inc. v. Cherry Auction, Inc., 76 F.3d 259, 262, 264 (9th Cir. 1996), and paraphrased by the U.S. Supreme Court. See Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, Ltd., 545 U.S. 913, 930 (2005) (“One infringes contributorily by intentionally inducing or encouraging direct infringement, and infringes vicariously by profiting from direct infringement while declining to exercise a right to stop or limit it.”) (citations omitted).

The line between direct liability and secondary liability is potentially blurry in the Internet age because all sorts of computers are involved in copying and transmitting pieces of a work from one location to another. The leading case is from the outset of the Internet age. In Religious Technology Center v. Netcom On-Line Comm. Servs., 907 F. Supp. 1361 (N.D. Cal. 1995), a user posted allegedly infringing materials to an electronic bulletin-board service that automatically copied and forwarded all of the posts to other users. The plaintiff copyright owner sued the user, the bulletin-board service, and the user’s Internet service provider (Netcom) for infringement. U.S. District Judge Ronald Whyte of the Northern District of California refused to hold Netcom liable for the copying done by a user:

The court believes that Netcom’s act of designing or implementing a system that automatically and uniformly creates temporary copies of all data sent through it is not unlike that of the owner of a copying machine who lets the public make copies with it. Although some of the people using the machine may directly infringe copyrights, courts analyze the machine owner’s liability under the rubric of contributory infringement, not direct infringement.

907 F. Supp. at 1359. More succinctly, “[a]lthough copyright is a strict liability statute, there should still be some element of volition or causation which is lacking where a defendant’s system is merely used to create a copy by a third party.” 907 F. Supp. at 1370.

Judge Whyte’s ruling has been adopted in all circuits that have considered the issue. As a result, a rich body of case law has developed distinguishing direct liability from secondary liability. That case law is not always consistent; but all courts agree that there cannot be secondary liability of any kind (contributory or vicarious) unless direct infringement has occurred.

Legal Background: The “Server” Test

The Ninth Circuit applied these two lines of authority to the public display right in Perfect 10 v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir. 2007). Perfect 10, a purveyor of soft-core pornography, sued Google (among others) for allegedly infringing its photographs of nude models. The photos had been copied by third-party infringers and posted online, where they were indexed in Google’s Image Search engine.  At the time, Google Image Search functioned by searching the Internet for images, and copying and storing low-resolution “thumbnail” versions of the images on its servers, together with the text accompanying those images. When a user searched for images using a text prompt, Google would retrieve and display relevant thumbnail images to the user. When a user clicked on a thumbnail, they would see the full-size infringing image on their computer. But Google did not store full-size copies of the infringing images on its servers; instead, “Google [stores and] provides HTML instructions that direct a user’s browser to a website publisher’s computer that stores the full-size photographic image.” 508 F.3d at 1161. That process was referred to as “in-line linking” in the case, 508 F.3d at 1156, but today it is more commonly known as “embedding.” Hunley, slip op. at 18-19.

With regard to the thumbnail images, the Ninth Circuit held that Google publicly displayed those images:

[B]ased on the plain language of the statute, a person displays a photographic image by using a computer to fill a computer screen with a copy of the photographic image fixed in the computer’s memory. There is no dispute that Google’s computers store thumbnail versions of Perfect 10’s copyrighted images and communicate copies of those thumbnails to Google’s users. Therefore, Perfect 10 has made a prima facie case that Google’s communication of its stored thumbnail images directly infringes Perfect 10’s display right.

508 F.3d at 1160. (The Ninth Circuit ultimately ruled, however, that making and displaying thumbnail images to facilitate an image search engine was a fair use. 508 F.3d at 1168.) But the Ninth Circuit also held that Google was not directly liable for publicly displaying the full-size infringing images:

Because Google’s computers do not store the [full-size] photographic images, Google does not have a copy of the images for purposes of the Copyright Act. In other words, Google does not have any “material objects … in which a work is fixed … and from which the work can be perceived, reproduced, or otherwise communicated” and thus cannot communicate a copy.

Google [stores and] 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.

508 F.3d at 1160-61. This test for distinguishing direct infringement from secondary liability has become known as the “server” test.

Note that the “server” test only absolved Google of direct liability for infringement. The Ninth Circuit further held that Google could still be liable for contributory infringement for linking to full-size infringing images, “if it had knowledge that infringing Perfect 10 images were available using its search engine, could take simple measures to prevent further damage to Perfect 10’s copyrighted works, and failed to take such steps.” 508 F.3d at 1172. This holding essentially means that a website operator that implements a “notice-and-takedown” system for user-generated content will not be held liable for contributory infringement in the Ninth Circuit.

Critiques of the “Server” Test

The Perfect 10 case undoubtedly reached the correct result on its facts. Under Perfect 10, if a website operator has actual knowledge that a specific image posted by a third-party is infringing, it must remove or disable access to that image, or it may be held liable for contributory infringement. Absent such knowledge, however, a website operator should not be held liable for publicly displaying infringing images that were posted to the Internet by others. An image search engine simply could not function if it could be held directly liable for publicly displaying all of the infringing images that have been made available on the Internet.

But was the problem in the Perfect 10 case that Google did not possess a “copy” of the infringing images in its possession, or that it did not act with the necessary “volition”? Recall that all of Google’s allegedly infringing conduct occurred automatically, as the result of an automated system that copied all images publicly available on the Internet, and generated thumbnails and embedded in-line links to those images automatically. This fits comfortably within Netcom’s holding that the “act of designing or implementing a system that automatically and uniformly [displays] copies of all data sent through it” should be analyzed under the rubric of contributory infringement rather than direct liability. The Ninth Circuit could have reached the exact same result in the Perfect 10 case by relying on Netcom rather than by adopting the “server” test.

The leading decision to break ranks with the Ninth Circuit is Goldman v. Breitbart News Network, LLC, 302 F. Supp. 3d 585 (S.D.N.Y. 2018). In Goldman, the plaintiff’s photograph was reproduced (without authorization) in several tweets on Twitter. Plaintiff then sued various news outlets that “embedded” the tweets in their news stories by including HTML code that would cause the tweet to be dis­played at a particular point in the article. The district court rejected the “server” test: “Nowhere does the Copyright Act suggest that possession of an image is necessary in order to display it.” 302 F. Supp. 3d at 593. [To be clear, the plain language of the Copyright Act requires that the defendant must display a “copy” of the image; but it does not say that the “copy” must necessarily be in the defendant’s possession.] Moreover, it held that even if the “server” test was appropriate in a case like Perfect 10, where the user had to click on a link to display the full-size image, it did not apply in a case where the embedded image was displayed automatically to the user, without the user taking any additional action. Id. at 595–96. Instead, the news outlets were (potentially) liable for taking deliberate, volitional action to cause the tweets to be displayed. [Eric’s note: the Goldman v. Breitbart case settled before definitively resolving the issues.]

A handful of other district courts have agreed with Goldman in rejecting the “server” test in the context of deliberate “embedding” of allegedly infringing content. See Nicklen v. Sinclair Broadcast Group, Inc., 551 F. Supp. 3d 188, 195 (S.D.N.Y. 2021); McGucken v. Newsweek, LLC, 2022 WL 836786, *5-6 (S.D.N.Y. Mar. 21, 2022); Leader’s Institute, LLC v. Jackson, 2017 WL 5629514, *11 (N.D. Tex. Nov. 22, 2017).

Hunley Reaffirms the “Server” Test

Applying Perfect 10, the district court in Hunley dismissed the action. Under the “server test,” the only “fixed copy” of the Instagram post resides on Instagram’s servers, so it is Instagram that is publicly displaying the photos, not Time or BuzzFeed. Instagram, of course, was licensed to publicly display the post under its own terms and conditions, so there is no direct infringement.  Instagram does not claim that their license covers any third parties that embed the content; but without any direct infringement, there cannot be any secondary liability.

On appeal in Hunley, the Ninth Circuit considered and rejected four arguments for limiting or overruling Perfect 10. First, the plaintiffs argued that “the Server Test should apply only ‘to search engines or other automated, algorithmic indexing platforms’ and should not extend to ‘content embedded into commercial websites from social media platforms.’” Slip op. at 19-20. The panel responded:

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.  [Slip op. at 23; emphasis in original]

Second, the plaintiffs argued that the server test was inconsistent with the text of the Copyright Act and the legislative history. The panel responded with a procedural objection:

We will not consider these arguments in any detail because they are foreclosed by Perfect 10…. 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. [Slip op. at 24]

Third, the plaintiffs argued that the server test was inconsistent with the subsequent Supreme Court decision in American Broadcasting Company v. Aereo, 573 U.S. 431 (2014). The panel disagreed for two reasons: first, Aereo involved the public performance right, which is not defined in terms of performing a “copy” of the work; whereas Perfect 10 involved the public display right, which is defined in terms of displaying a “copy” of the work. [Slip op. at 27-28] Second, it thought that Perfect 10 was not “clearly irreconcilable” with Aereo, and in particular, with Aereo’s comments regarding the perception of users. [Slip op. at 30-32]

Finally, “Hunley, Instagram, and their amici have peppered us with policy reasons to uphold or overturn the Server Test.” [Slip op. at 33] Again, the panel responded with the procedural objection: “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.” [Id.]

The consequence of upholding the server test was a foregone conclusion: under the server test, neither Time nor BuzzFeed were liable for direct infringement, because embedding HTML code is not a “public display” of the content found at the online location. Without direct infringement, Instagram could not be secondarily liable, and the Ninth Circuit affirmed the dismissal.

Prof. Ochoa’s Commentary

Note that the photographers could easily have stopped the alleged infringement by simply deleting their original Instagram posts. The ruling in Hunley essentially forces the photographers into an all-or-nothing choice: allow all others to link to and “embed” their Instagram posts without further consent or payment; or remove their Instagram posts in their entirety. In this scenario, I have little sympathy for the photographers; they knew what the consequences would be if they posted their own photos to Instagram.

However, there are other situations in which I would be much more sympathetic to them. For example, suppose a photographer licenses one online publication to publicly display a work in exchange for money. Does it make sense to then allow every other online publication to “embed” the work without further consent or payment? Surely paid licensing should not have to work on an all-or-nothing basis. If we allow embedding in this scenario, free-riding essentially destroys the financial incentive of the first publisher to pay for the photos (unless it hides them behind a paywall), and it essentially destroys the financial incentive for the photographer to take and sell the images in the first place.

[Eric’s comment: in this scenario, embedding a photo from a third-party site necessarily runs the risk that link rot will break the display, which is a lousy experience for users. This risk deters many potential users from doing third-party embedding. Furthermore, the site hosting the embedded photo can affirmatively discourage or prevent embedding, such as displaying prominent watermarks and implementing technological limits on linking, and the photographer can require the publisher to take those steps as a condition of the copyright license if they care.]

The Ninth Circuit is, of course, correct that public display of a “copy” of a work requires that there be a “copy” that is “fixed in a tangible medium of expression.” In my opinion, however, nothing in the Copyright Act requires that the defendant be in possession or control of that copy in order to be held directly liable for public display. Instead, the better-reasoned precedent only requires that the defendant engage in “volitional conduct” that directly causes the public display to occur. In other words, the line between direct liability and contributory liability should be a distinction between automated copying that merely enables the public display of a work (when the user has to take additional action, as in Perfect 10) and intentional, deliberate “embedding” of a work that causes the work to be displayed without further action by the user.

In Perfect 10, the “embedded links” to full-size images were not displayed automatically; it required two further actions by the user to cause the images to be displayed.  First, the user had to search for images associated with particular text; and second, the user had to click on the thumbnails in order to display the full-size infringing images. In that context, the distinction between direct liability and contributory liability makes sense: Google did not post the images, and it did not cause them to be publicly displayed; it merely helped people find images that were already publicly made available by others.  In Hunley, however, the alleged public display was not automated; the authors of the articles involved made a deliberate decision to “embed” an image that had been posted to a different social media platform. That is the volitional conduct that satisfies direct liability. That the alleged direct infringers (BuzzFeed and Time) did not have possession or control of the “copy” should not change the fact that they should (potentially) be directly liable (assuming they acted without authorization).

This does not mean that the Ninth Circuit reached the wrong result. Hunley did not sue Time or BuzzFeed; she sued Instagram for making “embedding” available to others. But Instagram did not induce or cause the alleged direct infringers to embed the images involved; Instagram simply made a tool available that was capable of substantial non-infringing use. The vast majority of people who post to social media want their posts to be embedded and made available to others. It is very “troll-like” behavior for Hunley to post images to Instagram and then profess to be “shocked, shocked” that others have embedded those posts. The problem arises because copyright law is written as an “opt-in” system (no copying without authorization), while the Internet is designed as an “opt-out” system (it allows copying unless it is prohibited). Given that characteristic, a copyright owner who posts their own work ought to have to take additional steps to prevent or discourage copying before it can claim that their work was infringed by others. In either case, Instagram certainly should not be held liable for infringement committed by others.

Of course, the line between direct and contributory infringement isn’t expressly found in the text of the Copyright Act, so ultimately the question is really one of public policy. In terms of policy, the “server” test has some distinct advantages. First, it is easy to apply: it is easy for the parties to determine where the allegedly infringing image is being stored (just look at the HTML code), and it relatively easy to determine who has possession or control of that website. Second, it fits better with the “opt-out” design of the Internet, so it makes it easier to dispose of “troll-like” cases like the Hunley lawsuit. But it also has distinct disadvantages: First, it makes it difficult for copyright owners to selectively license their images to platforms that are willing to license them, because free-riders can simply wait for the first authorized image to be posted, and they can then “embed” that image without having to pay for it. How are professional photographers supposed to make a living if commercial media outlets can “embed” their images without paying for them? Second, as explained above, the “server test” essentially forces professional photographers to choose between putting their own images online and letting everyone link to those images for free, or trying to make a living by not licensing their photos for online use at all except behind a paywall.

The “server test” also sometimes results in liability for parties that took no volitional action whatsoever. In Bell v. Wilmott Storage Services, 12 F.4th 1065 (9th Cir. 2021), for example, the defendant purchased a website from another party. There was an infringing image on its server, but there were no active links to that image. Nonetheless, the Ninth Circuit held that the defendant had violated the public display right: “Although a member of the public could not access the photo by simply visiting Wilmott’s website, applying the server test, Wilmott’s server was continuously transmitting the image to those who used the specific pinpoint address or were conducting reverse image searches using the same or similar photo. … By displaying the [infringing] photo on a server that was publicly accessible to anyone with an Internet connection, … Wilmott publicly displayed the photo, regardless of whether or not any particular person actually found and viewed it.” 12 F.4th at 1073.  This is an absurd result; the notion that any member of the public had actually viewed the image was near zero. Thus, the server test is both underinclusive and overinclusive.

Ultimately, the problem is that copyright law compensates photographers and other copyright owners through a model of artificial scarcity. Society needs to find a way to compensate such copyright owners without encouraging paywalls and other means of limiting access to copyrighted works. One alternative is some sort of compulsory license: anyone who wants to embed images has to pay a fee to a collecting society, which will distribute the money to copyright owners. But the administrative costs of running such a system are likely to be prohibitive; and Americans are not likely to accept the type of government-run collecting societies that exist in other countries.  (One can hear the protests of “that’s socialism!” already.) The Copyright Claims Board, an administrative court for resolving “small claims” in copyright law (less than $30,000), may yet prove to be useful; but so far, it has generated a lot of litigation “churn” without much in the way of substantive relief.

I don’t claim to have the answer to this dilemma; and regardless, courts have to resolve disputes with the law that currently exists. Hunley deepens the split between the Ninth Circuit and district courts in other jurisdictions. I find the “server test” unsatisfying because it enables actors to deliberately use copyright images while easily avoiding payment or liability for infringement. But those who like the “server test” should hope that the Ninth Circuit takes this case en banc, because the facts are much more favorable to the defendants than in most such cases.

Eric’s Additional Comment

I note the overlap between the litigation over embedding and the efforts, such as the California Journalism Protection Act (CJPA), to impose “link taxes” on some Internet services for “displaying, providing, distributing, or directing users to” third-party “journalism.” In practice, the CJPA would impose payment obligations on some embedders for some works, even though this case would hold that those uses are not infringing. I think this ruling reinforces why the CJPA is preempted by federal copyright law. CJPA tries to impose copyright law-like protection for situations that federal copyright law squarely addresses. Either way, you should pay close attention to efforts like CJPA that attempt to regulate ubiquitous Internet practices using a non-copyright legal framework.

Conclusion

Is the world safe for embedding after Hunley? In cases litigated within the Ninth Circuit, yes; and many cases are filed in the district courts of the Ninth Circuit because that is where many tech companies are based. But in the district courts of the Second Circuit, and possibly elsewhere, the answer is still no.  Indeed, the split of authority has already affected reporting on social media; most companies have decided to be “risk averse” and to stop embedding social media posts in their articles, in favor of hyperlinks that the user has to click to see the original content. This is bad for readers, who now have to take an active step to inform themselves about the subject of such articles. It is also bad for the copyright ecosystem. See James Gibson, Risk Aversion and Rights Accretion in Intellectual Property Law, 116 Yale L.J. 882 (2006). Finally, this split of authority is rife with opportunities for forum shopping, and one suspects there will be a great deal of money spent litigating the ancillary issues of personal jurisdiction and venue, because the consequences of litigating in one jurisdiction or the other are so drastically different.

Three times in the opinion, the panel mentioned that if Hunley thought Perfect 10 was wrong, “the proper procedure is to seek rehearing en banc.” [Slip op. at 25; see also slip op. at 24, 33] One assumes the plaintiffs will take the suggestion and file a petition for rehearing en banc. If that fails, the plaintiffs may try a petition for certiorari to resolve the split in authority. Either way, it seems highly likely that we have not yet heard the last of this case.

Update (June 2024)

In Miller v. 4Internet, LLC, No. 22-16195 (9th Cir. June 28, 2024), the court said:

On May 1, 2024, the Hunley panel voted to deny the petition for rehearing en banc….

Under the “server test,” a website publisher is only liable for direct infringement if the publisher stores the infringing image on its own server rather than embedding or linking from a third-party server….Miller originally contended that the server test is (1) inconsistent with the language and intent of the Copyright Act and (2) no longer good law after American Broadcasting Cos. v. Aereo, Inc., 573 U.S. 431 (2014). Miller also argued that even if the server test was still valid, it did not foreclose his claim for vicarious liability. After the parties in this case submitted their briefs, our circuit considered and rejected identical arguments in a different case, Hunley v. Instagram, LLC, 73 F.4th 1060 (9th Cir. 2023).

UPDATE 2: In Great Bowery v. Best Little Sites, 2024 WL 3416038 (D. Utah July 15, 2024), a Utah district court judge went out of his way to disagree with the Hunley v. Instagram ruling, essentially begging the 10th Circuit to reignite the legal battles over embedding. I’m praying that doesn’t come to pass.

UPDATE 3: In McGucken v. Valnet, Inc., No. 24-511 (9th Cir. 2024), the Ninth Circuit reaffirmed the Hunley server test:

Because Valnet’s website contained only embedded references to McGucken’s Instagram posts, Valnet never displayed or distributed copies of the works to anyone on the internet. Thus, under the Server Test Valnet did not infringe McGucken’s display rights.

McGucken suggests that the Server Test “should be cabined to use in the search-engine context.” This argument was considered and rejected in Hunley. As we explained, “in articulating the Server Test, Perfect 10 did not rely on the unique context of a search engine” but instead the method of embedding an image.