Sony Takes Action Against Deepfake Fraud: Removes 135,000 Songs Impersonating Artists

Sony Music says it has requested the removal of more than 135,000 songs generated by fraudsters using artificial intelligence to impersonate its artists on streaming services. These so-called deepfakes target some of Sony's major acts, including Beyoncé, Queen, and Harry Styles.

The presence of these counterfeits leads to direct commercial harm to legitimate recording artists, as they often emerge during periods when these artists are promoting new albums. Sony's President of global digital business, Dennis Kooker, remarked that in the most severe cases, these deepfakes risk damaging a release campaign or tarnishing an artist's reputation.

As artificial intelligence technology becomes cheaper and easier to access, the company has noted that the number of deepfake songs is increasing. They believe the 135,000 tracks they have discovered represent just a small portion of the total fraudulent entries on streaming services. Since last March, Sony has identified approximately 60,000 songs falsely claiming to feature its roster of artists, with other affected musicians including Bad Bunny and Miley Cyrus.

Kooker explained that these fake tracks exploit moments when artists are in the spotlight, thereby taking advantage of the demand generated by their promotional efforts. This situation is compounded by streaming fraud, which involves fake artists uploading music to platforms like Spotify and YouTube in hopes of increasing their play counts and reaping royalty payments.

According to the International Federation of the Phonographic Industry (IFPI), up to 10% of content across all streaming platforms may be fraudulent. As AI continues to proliferate in the industry, there are growing calls for regulations and transparency to protect authentic artistic creation and user experiences.

The music industry has noted that the ongoing challenges around AI-generated music and streaming fraud highlight the need for effective tools to identify and label such content, ensuring that fans can distinguish between real artistry and unauthorized AI creations. Without proper identification, the trust and transparency crucial for a sustainable music ecosystem are at risk.