Last month, music streaming giant Spotify patented ‘Plagiarism Risk Detector and Interface’ technology, utilising artificial intelligence to compare the melody, lyrics and harmonies of new uploaded music to those pre-existing on their database to generate a similarity value and detect plagiarism.
The company currently screens for plagiarism manually, using experts and lawyers alike. However, they note that this new technological movement will not only be more practical, but also useful in helping artists in detecting similarity during composition. Furthermore, Spotify hopes for a much higher success rate in detection, as the software algorithm aims to screen and compare ‘vast collections’ of material in a shorter space of time. Although the technology has not yet been fully developed, Spotify posits it will also be able to detect the specific part of any proposed song that may be in violation of another artists’ copyright, as well as detect fragments which are completely original.
The patent has already faced some criticism, as speculation surrounds how the software will deal with common chord progressions and melodies. Questions as to Spotify’s motivation have also been raised, as the patent may be considered to protect the service from litigation, as opposed to protecting the works and rights of artists.
Although it is unclear as to the exact progress of development, concerns have arisen as to the practicality of the software; YouTube currently operate a similar copyright detector for videos uploaded to its service, however, there have been many instances where this has proven to be flawed, falsely accusing users of copyright infringement.
The music streaming service has responded stating that not all patents registered are eventually incorporated into their product and, furthermore, have refused to confirm whether the technology has yet been implemented.
By Ellie King, student from Southampton Solent University
