Musicbusinessworldwide.com
The legal battle between major record labels and AI music platform Suno has entered a new phase, with Universal Music Group and Sony Music Entertainment asking a federal court to deny the company’s request to keep key training-data information out of public view.
In a filing submitted to the U.S. District Court for the District of Massachusetts, the labels argued that Suno has failed to justify sealing the total number of audio files used to train its generative AI music system. According to the companies, the figure is directly relevant to the core copyright questions at the center of the case. It should remain accessible as part of the public court record.
The disputed statistic, referred to in court documents as the “Model Training Figure,” represents the overall volume of recordings that the labels allege were incorporated into Suno’s training dataset. While the number was initially redacted from recent filings, Universal and Sony said they reserved the right to challenge any attempt to keep the information permanently confidential.
Suno has maintained that revealing the figure could expose sensitive business intelligence, potentially allowing competitors to conclude the company’s development process and technical strategy. The startup argues that such information could provide rivals with an unfair advantage in the increasingly competitive generative AI marketplace.
The labels, however, contend that those concerns are speculative and insufficient to outweigh the public’s right to access information contained in court proceedings. They further noted that Suno has already publicly acknowledged training its technology on tens of millions of recordings, arguing that disclosure of a more precise number would not meaningfully reveal proprietary information.
The dispute unfolds alongside a broader effort by Universal and Sony to expand their lawsuit. Last month, the companies sought permission to add more than 61,000 additional recordings to the case after allegedly identifying their works within Suno’s training data using audio-fingerprinting technology. Suno has opposed the expansion, claiming it would delay resolution of key legal questions, including its fair-use defense.
About the Author
