chevron-down Created with Sketch Beta.
June 07, 2024 2024 ELI Writing Competition

Runner-Up Essay: A “Perfect Storm” for Reworking the Copyright Test

Emily Cohen


Katy Perry and Juicy J’s “Dark Horse” debuted at number 17 on the Billboard Hot 100 in 2013 and reached number one in 2014, remaining on the list for 57 weeks. The hit was pre-dated by “Joyful Noise” by FLAME, a 2009 Christian rap song with 1,365,041 total YouTube views and 2,465,724 plays across two Myspace pages at roughly the time of Perry’s recording. “Joyful Noise” never made the Billboard Hot 100. In 2014, FLAME filed a lawsuit, alleging that the defendants’ song “Dark Horse” infringed upon the plaintiffs’ copyright in “Joyful Noise.”

All defendant songwriters claimed they did not listen to “Joyful Noise” before writing “Dark Horse.” Therefore, FLAME needed circumstantial evidence of access, which courts have interpreted as “a reasonable possibility, not merely a bare possibility, that an alleged infringer had the chance to view the protected work.” But what is a “reasonable possibility” in the post-internet era? While courts have found “the availability of a copyrighted work on the Internet in and of itself, is insufficient to show access through widespread dissemination,” it is unclear when a work is sufficiently available online to be accessible. In this case, the defendants could have “had a chance” to hear the song through a Google search, but this is true for almost every song today. Therefore, the relevant question is what is the dividing line between a song being sufficiently and insufficiently accessible for a reasonable possibility of access?

In the modern era, any person can gain access to streaming platforms and then theoretically will also have access to any song posted on these platforms. Therefore, the copyright test, with its reliance on whether a defendant could have accessed a song, no longer functions properly.

I. The Current Copyright Doctrine

Musicians suing for copyright infringement must prove: (1) they have a valid copyright and (2) the defendant copied some protected element of their song. Valid copyright ownership is usually not disputed, as it is relatively easy to prove. Instead, infringement suits in music focus on the second part: proving the defendant copied protected elements in the plaintiff’s song, which breaks down into a basic two-part test.

1. Copying.After establishing the musician has a valid copyright in his song, the plaintiff must prove the defendant copied the plaintiff’s song. Ideally, a plaintiff could provide direct evidence that the defendant copied his song; however, because this is highly unlikely, the plaintiff “can attempt to prove [actual copying] circumstantially by showing that the defendant had access to the plaintiff’s work and that the two work share similarities probative of copying.” And “[t]o prove access, a plaintiff must show a reasonable possibility, not merely a bare possibility, that an alleged infringer had the chance to view the protected work.” This can be done by (1) showing the plaintiff’s song is linked to the defendant’s access through a series of events, or (2) proving the plaintiff’s song has been “widely disseminated,” and thus is accessible by the defendants.

2. Unlawful Appropriation.Once a plaintiff establishes that the defendant actually copied the plaintiff’s song, the plaintiff must prove the defendant unlawfully appropriated the plaintiff’s song by copying in a way copyright law deems inappropriate. The key question is whether the two songs share substantial similarities. The bar for finding similarities probative of copying only requires some evidence of similarities that makes one suspicious that the songs were not independently created.

II. Why the Access Prong is Outdated

Since the test for copyright infringement was first created, the way music is created and disseminated has evolved significantly. Therefore, the current copyright test, particularly the access prong, does not align with the current state of music.

A. How Society Consumes Music Has Changed

Previously, the main way people listened to music was through radio, cassette tapes, CDs, and records. However, as of 2021, only 31 percent of Americans surveyed most commonly consume music through radio and seven percent through CDs, cassettes, or vinyl records.

The most common way to consume music is through streaming services, which is how 41% of Americans surveyed consume music. Spotify has 205 million paying subscribers as of 2022 while Apple Music has 78 million paying subscribers as of 2021. Any song on a streaming service is available to users simply by searching or by discovering it through curated playlists created by the service for the individual user. Accordingly, while an individual may never have stumbled across an artist before the rise of streaming platforms, these platforms make it easier to discover smaller artists in multiple genres.

With this new way of consuming music, technically any song on a streaming platform or a social media site is accessible to anyone who uses that platform. However, with the large number of songs in existence, what is the likelihood one will actually access a given song? Accordingly, the definition of access as it exists in the copyright test seems unworkable given the modern access to millions of songs at any moment.

B. Society Listens to More Music Daily

In large part because of the shift in how people consume music, the average person listens to more music than before. As of 2022, the average person listens to 20.1 hours of music per week. This is an increase from 18.4 hours in 2021 and 18 hours in 2019.

Another reason for the increase is the prevalence of music on social media platforms like Instagram, Facebook, and, most recently, TikTok. Eight percent of people surveyed said they use short-form video-sharing platforms like TikTok as their primary source for accessing music. TikTok has seen a major increase in users over the past few years, with 150 million Americans on TikTok as of March 2023. This platform has been used by young artists to garner traction for their songs. Users like to find new music on TikTok, with 63 percent of people surveyed claiming music is central to their time on the platform. Because videos on TikTok are short, if one spends just one hour on TikTok, he could potentially listen to hundreds of songs.

Because of this increase in listening to music, society has access to and contact with more songs. With this increased access, it becomes more common that someone has heard a song without consciously knowing it, and courts have already found that subconsciously copying a song is actionable. Therefore, it becomes more likely that a defendant has a reasonable possibility of accessing or has accessed a song.

C. The Abundance of Songs

The shift towards streaming services and social media as spaces to access music has made it easier for individuals to release music. New songs are constantly added to streaming services, with almost 100,000 new songs added to Spotify each day, and anyone can upload music to streaming services. With more music, as Judge Learned Hand said, “[w]hile there are an enormous number of possible permutations of the musical notes of the scale, only a few are pleasing; and much fewer still suit the infantile demands of the popular ear. Recurrence is not therefore an inevitable badge of plagiarism.” As more songs are released, the likelihood each song will be completely novel decreases substantially. Accordingly, the copyright test must adapt to better fit the modern era characterized by increased access to music and the release of more songs, such that the possibility of creating a completely unique song is nearly impossible.

III. Consequences of Leaving the Test As Is

Because of the many changes in the music industry, keeping the current test for copyright infringement will have dire consequences for artists and the music industry.

First, the current copyright test punishes artists who use common patterns of notes that sound similar to earlier released songs. With the Perry case, the songs “Joyful Noise” and “Dark Horse” are, in the big picture, extremely different songs: “They are in different key signatures, have different chord progressions, and mostly use different instruments and tones.” The songs both have “a high-pitched synth-like instrument playing a similar melody,” and, thus, sound similar. With finite choices for creating well-received music, it is essentially inevitable that songs will use similar melodies or rhythms and accordingly sound somewhat similar.

Second, the current music copyright landscape opens the potential for an increase in copyright infringement suits. In the past, bringing a successful music copyright infringement suit was a relatively difficult task. However, recently, plaintiffs have been successful in lawsuits against well-known artists. There was a 128% increase in copyright cases filed from 2010 to 2019 and a 57% increase from 2018 to 2019. The idea that access can be established via a moderate number of streams or views combined with the relatively low bar for proving similarities probative of copying between two songs will encourage lawsuits.

Finally, the increased possibility of litigation for copyright infringement in music may stifle creativity and deter artists from releasing music. For example, Busbee, a songwriter for singers such as Keith Urban and Lady Antebellum, in response to a ruling against Robin Thicke and Pharrell Williams for “Blurred Lines,” claimed, “. . . it puts a massive damper on the [songwriting] process, if you’re concerned that you will be sued.” Copyright law needs a new test that meets the balance of protecting the artist but also allowing artists to be inspired by and build off of the works of others.

IV. Solution

Courts have not clearly addressed what exactly access means in the age of music streaming. What courts have said is that existence on the internet is not enough to establish access. Similarly, scholars have argued for a new definition of access that creates a heightened standard for what counts as sufficiently accessible.

However, instead of redefining access and creating a heightened definition of what is sufficient to establish access, there should be a rebuttable presumption of access in music copyright infringement lawsuits. The heightened standard should instead be applied, not to the access prong, but to the unlawful appropriation prong so copyright infringement is only found when there are substantive, prolonged similarities between two songs. This can be accomplished by applying a similar “likelihood of confusion” test from trademark law to music copyright law, like an idea proposed by David Nimmer. Because the heightened unlawful appropriation test will inevitably require there to be similarities probative of copying between the two songs, this part of the actual copying prong is not included in the proposed new test.


Within the modern landscape, it is time for music copyright law to evolve and afford greater protections to artists, such that copyright law does not stifle creativity. Therefore, courts should change the test for copyright infringement in music to include two modified parts: (1) a rebuttable presumption of access, and (2) unlawful appropriation due to a “likelihood of confusion” of the two songs in the marketplace.

    The material in all ABA publications is copyrighted and may be reprinted by permission only. Request reprint permission here.

    Emily Cohen

    Duke University School of Law

    At the time of writing, Emily Cohen was a Third Year Law Student at Duke University School of Law, Durham, NC, and was a candidate for a JD in May 2024.