Last year, record companies and their American representative, the RIAA (Recording Industry Association of America), filed a lawsuit against Udio and Suno. These two startups allow their users to generate music from text queries. But of course, these synthetic tracks don't come out of nowhere: the startups' AI models were trained using a huge volume of existing songs. And they didn't ask for permission.
AI, a future goldmine for record companies?
The recording industry is therefore accusing Udio and Suno of copyright infringement, and the RIAA is seeking up to $150,000 for each infringement, a potentially painful sum of several billion dollars. An impossible bill to pay for the two targeted companies, whose defense—like the entire AI industry—is based on "fair use," which would give them the right to train their models without the consent of rights holders or financial compensation.
In parallel with these legal actions, a race against time has begun between the record companies (Universal, Warner Music, and Sony) and the startups to seal usage agreements. It's complicated, as Bloomberg explains: on the one hand, the labels want more control over the use of their productions, on the other, Udio and Suno are looking for a certain flexibility to experiment... and not to pay too much.
If an agreement were to be reached, it would mark the end of the legal proceedings, and above all, it would represent a framework paving the way for a profitable coexistence between generative AI and the music industry. A precedent that could extend to other creative sectors grappling with the same issues.
Source: Bloomberg
0 Comments