Sony Claims They Can Detect the Use of Protected Music in AI Generated Songs

Sony says it may have found a way to solve one of the music industry’s biggest modern problems. As artificial intelligence floods streaming platforms with generated songs, identifying where those sounds come from has become increasingly difficult. The company now claims its technology can actually trace copyrighted material hidden inside AI-created tracks.

The newly revealed system comes from Sony’s broader AI division and is designed to analyze machine outputs across media, including music. Rather than simply flagging song as suspicious, the technology reportedly compares an AI-generated track against existing recordings to identify which copyrighted works were used. According to DigitalMusicNews, the tool can even estimate percentage contributions from specific human creators whose music may have been included in the training data or output. 

That level of detail could dramatically change the legal landscape around generative music. If accurate, it would allow rights holders to show not only that their material influenced an AI song but how much. Sony’s system can calculate contributions when developers cooperate by analyzing training data. If they do not cooperate, the tool instead compares the generated audio directly to existing songs to identify the original work. 

The announcement arrives during ongoing legal battles between record labels and AI music platforms. Sony Music is already involved in lawsuits against certain AI developers, and a reliable detection system could strengthen copyright claims and support royalty payments. It could also establish a pathway for compensation by identifying derivative use of protected recordings.

Sony is not alone in pursuing this technology, but its focus on precise attribution stands out. The company appears to be positioning the tool as a bridge between innovation and copyright protection, allowing artificial intelligence to continue developing while still preserving ownership rights. If the detection works as promised, it may reshape how AI music is licensed, monetized and regulated across the entire industry.