A research team led by the University of Texas has come up with a model called Ambient Diffusion which they claim “gets around” the issue of copyright and AI image generators by feeding the model images that have pixels missing — in some cases as much as 93%. “Early efforts suggest the framework is able to continue to generate high-quality samples without ever seeing anything that’s recognizable as the original source images,” reads a press release.
Source: AI Image Generator Avoids Copyright Issues by Training on Corrupted Photos
San Francisco-based blockchain-powered music platform Audius has partnered with rights administration platform Music Reports to power the licensing of rights and payments to music publishers through its Songdex Marketplace opt-in licensing platform. The new partnership follows Audius’ launch of its music marketplace last year, which provides a platform for its 6 million users to support artists with direct payments.
Company reports to have distributed more than €1bn to rightsholders in a consecutive 12-month period for the first time. With this distribution milestone passed,
WIN’s AI manifesto comes as music companies around the world navigate the complex new world of AI, working to maximize the technology’s potential for artists and businesses while addressing challenges such as deepfakes of artists and the unauthorized use of copyrighted music in training AI models. “The global independent music community welcomes new technological developments which respect the value of music and creators’ rights,” WIN CEO Noemi Planas said.


