Google and Universal Music are negotiating royalties to "rights holders" for content generated by machine learning systems. The negotiations come amid a surge in the publication of fictitious AI-generated music tracks that mimic well-known tunes or voices of performers. The authors of such imitations are expected to be able to distribute them legally by paying royalties, similar to the royalties paid for the use of original compositions.
This initiative may have consequences in the form of consideration of such practices by courts and legislators, as well as lead to the extension of copyright to AI-generated content, which is currently in the public domain (such distribution is possible, as programs and devices are not subjects of law, including such branch of law as copyright).
The benefits of the regulatory approach for large owners of AI systems may outweigh the costs associated with contracting with "rights holders" and paying them royalties, as such content producers will have a monopoly position as the only generators of content with certain characteristics whose activities are legal. The royalty-based approach jeopardizes not only machine learning systems for voice cloning, but all generative AI in general, including ChatGPT and the free models of OpenLLama and
No comments yet