Music plagiarism: Do computers really judge copyright infringement cases better than human ears?

AUCKLAND, New Zealand — In a year marked by high-profile music copyright cases, such as Ed Sheeran’s successful defense against accusations of ripping off Marvin Gaye’s “Let’s Get It On,” and Pharrell Williams and Robin Thicke’s failed attempt to prove that “Blurred Lines” wasn’t a copy of Gaye’s “Got to Give It Up,” there are questions about the potential role of automated algorithms in resolving such disputes. Could algorithms bring objectivity and efficiency to music copyright infringement decisions, reducing the number, scale, and expense of court cases?

“It’s the largest study so far of how the best algorithms compare with humans in judging when music crosses the line into plagiarism,” says Dr. Patrick Savage, a musicologist and senior research fellow at the University of Auckland’s School of Psychology, in a university release. “It’s fair to say that algorithms won’t be taking over any time soon.”

Dr. Savage’s involvement in the field of music copyright disputes also includes contributing expert evidence for a court case involving Katy Perry.

In this study, 51 individuals were tasked with assessing 40 instances of alleged plagiarism in music compositions spanning from 1915 to 2018. Examples included a New Zealand National Party campaign advertisement from 2014, reminiscent of Eminem’s style, and ex-Beatle George Harrison’s “My Sweet Lord” from the 1970s. To evaluate the cases, the team used two of the most reputable publicly available music plagiarism detection tools, PMI and Musly.

creating music using artificial intelligence
creating music using artificial intelligence. (© YarikL –

The study found that the assessments made by participants matched court decisions in 83 percent of the cases (33 out of 40), compared to algorithms, which achieved a 75 percent accuracy rate (30 out of 40). However, one notable limitation of the study is its assumption that the court decisions under review were indeed correct.

“The `Blurred Lines’ case caused considerable controversy – and neither our study participants nor the algorithms strongly supported the legal decision – nor did many musicians, musicologists, lawyers, or judges, for that matter,” explains Dr. Savage.

One permanent constraint on the use of algorithms in copyright cases is the potential influence of non-musical factors.

“For example, regardless of how similar two songs are, there won’t be a breach of copyright if the allegedly plagiarizing composer can show that it would have been impossible for them to have heard the earlier song,” notes Dr. Savage.

While algorithms are unlikely to replace jury trials entirely, their objective assessments could become a valuable factor in these cases.

“For example, Spotify is already experimenting with a Plagiarism Risk Detector that might help artists automatically catch unintended similarities with existing works before they release new songs,” says Dr. Savage. “Future court cases might also be able to include graphs of how similar two songs are in relation to past cases to give judges and juries more objective data and context to aid their decisions.”

As music copyright lawsuits become increasingly common, Dr. Savage and his study co-authors argue in their paper that “unjustified music copyright lawsuits not only inhibit music creativity but also waste millions of taxpayer dollars annually to cover the adjudication of these disputes.”

The study is published in the journal Transactions of the International Society for Music Information Retrieval.

You might also be interested in:

YouTube video

Follow on Google News

About the Author

StudyFinds Staff

StudyFinds sets out to find new research that speaks to mass audiences — without all the scientific jargon. The stories we publish are digestible, summarized versions of research that are intended to inform the reader as well as stir civil, educated debate.

The contents of this website do not constitute advice and are provided for informational purposes only. See our full disclaimer