Music copyright is still a mess after 500 years. Can we finally change things before AI takes over? – National

Controlling the distribution of music — and thus making sure composers get paid for their labour and talent — has been a problem that dates back to the invention of the printing press.

In 1498, less than 50 years after Johannes Gutenberg revealed the printing press, a savvy entrepreneur named Ottaviano Petrucci received a patent from the Venetian Senate for publishing musical notation with one of these new-fangled machines, giving him a monopoly on sheet music. He controlled the copyright and publishing of all music. But then in 1516, Pope Leo X stripped away Petrucci’s power when it came to organ music and gave it all to Andrea Antico, someone who pleased the pontiff more.

This mess continued through the centuries. In England, Elizabeth I granted William Byrd and Tomas Tallis a patent on all music publishing, which not only included all music created in the kingdom but also prohibited foreign vendors from peddling their music in England. The cherry on top was that Byrd and Tallis also owned the rights to the printing of blank music paper. In other words, if you were an English composer, you had to pay them even before you wrote down a single note. Soon after, a French composer named Jean-Baptiste Lully managed to secure control over all operas performed in France and became one of the wealthiest people in the country.

It took a while for these royal-granted monopolies to be wiped out, leading to the Berne Convention of 1886, which set the first true international standards for who had the right to copy and distribute intellectual property with a focus on the rights of the creators and not the publishers. Those terms have been renegotiated a number of times in the last century-and-a-half. Meanwhile, technology marched on, adding new levels of complexity to protecting the rights of artists, especially in the digital age.

One area that’s blown up is allegations of copyright infringement by one musical artist upon another. We’ve seen it with cases involving George Harrison and the Chiffons, Marvin Gaye and both Robin Thicke and Ed Sheeran, Chuck Berry and the Beach Boys, Sam Smith and Tom Petty, Vanilla Ice vs. David Bowie and Queen, The Hollie and Radiohead, Spirit and Led Zeppelin, and dozens of others. These accusations of plagiarism — many completely unfounded, in my view — have sucked up an enormous amount of court time and money.

There’s a thriving industry of ambulance-chasing lawyers who “discover” that a newer song has certain sonic similarities to a song from the past. The composer of the older song is contacted and told that if they sign on, there could be a songwriting credit for them on the new song (meaning that they’ll get a stream or royalties) or at the very least receive some kind of out-of-court settlement. Dua Lipa is currently facing three such lawsuits, the latest being over an alleged unauthorized sample in her hit Levitating. It’s all very nutty, especially the current “dembow” case that seeks to upend the rhythmic foundations of music.

With so many competing interests, unclear statutes, differing interpretations between territories, gullible juries and advancing technology, protection of copyright is just as much a disaster as it was in the days of Petrucci and Antico.

Underpinning all this is a mathematical fact: There remain just 12 notes in the western scale and a finite number of ways they can be combined into pleasing combinations. With 100,000 new songs being uploaded to streaming music services every day, unexpected and unintentional duplication is inevitable. And with AI-composed music quickly being adopted, the situation will get even worse.

Or will it? Probably, but there have been some interesting developments of late.

First, Judge Beryl Howell of the U.S. District Court for the District of Columbia ruled that any kind of art — including music — solely created by artificial intelligence cannot be subject to copyright. Why? Because “human authorship is an essential part of a valid copyright claim.” This is in line with some rules followed in Canada. Meanwhile, the people in charge of the Grammy Awards have new guidelines that say “only human creators” can win an award. “A work with no human authorship is not eligible in any category.” That may be, but they haven’t ruled out considering songs that feature a portion created by AI, so we’ll call that half a win for humans.

But Damien Riehl and Noah Rubin want to settle this once and for all. They’ve created an algorithm that can generate 300,000 eight-note melodies every second in order to create a database of 68 billion “songs.” Those melodies were then copyrighted and released online into the public domain, meaning that they’re usable by anyone. They claim that these files — which sit on a small hard drive — contain “every melody that’s ever existed and ever can exist…. No song is new. Noah and I have exhausted the data set. Noah and I have made all the music to be able to allow future songwriters to make all of their music.”

Their point? That copyright law is completely broken and needs to be updated properly. Riehl outlined everything in a TEDx talk.

The Riehl/Rubin conjecture has yet to be tested in court, but it’s inevitable that it will be. I look forward to the outcome.

&copy 2023 Global News, a division of Corus Entertainment Inc.