Spotify is developing an AI to tell songwriters if they’re plagiarising
The tool appears to work in real-time and with a graphic interface not unlike your average notation software.
Image: Chesnot / Getty Images
What if you could predict crime and prevent it from happening before it’s committed? No – you aren’t reading a synopsis for the 2002 sci-fi thriller Minority Report starring Tom Cruise. Spotify appears to be developing AI technology to inform songwriters if they’re plagiarising music, so they can avoid hefty copyright lawsuits before they happen.
A recent filing at the European Patent Office has detailed Spotify’s new project, a “plagiarism risk detector and interface”. What’s interesting, is that the tool appears to be targeted at music-makers; it’s designed to work in real-time and with a graphic interface.
The technology deals entirely with lead sheets – i.e. notated melodies and chord information – to compare a song in progress against ones in a database. Not only will it let you know if you’re treading on another artist’s intellectual property, it also tells you the degree of your plagiarism.
According to the patent document, chord sequences, melodic fragments, harmony and other musical elements will all be taken into account. The GUI then lets you know if your idea has appeared in “many works,” “some works” or is “completely new”.
Spotify explained in the patent document, that the tool would “allow artists to generate lead sheets more quickly and confidently by detecting and providing visual feedback as to whether any aspect of the work has a probability of being deemed plagiaristic.”
The company added: “Software-assisted detection for text plagiarism on the other hand allows vast collections of documents to be compared to each other, making successful plagiarism detection much more likely”.
When, if and how Spotify plans to deploy this technology is up in the air for now. Besides that however, the tool could pose some serious questions in regards to the future of music copyright laws.
While the tool may be positioned as a means to defend artists from accidental plagiarism, it could, in theory be deployed as a means to measure plagiarism to serve as evidence in a court of law.
In other copyright news, back in February, Damien Riehl, a musician, programmer and copyright attorney, teamed up with programmer colleague Noah Rubin to algorithmically write every possible melody in existence to MIDI and make them open source to help musicians circumvent copyright laws.
Read the document in full here.
For more music technology news, click here.