Meta and UMG strengthen agreement against unauthorized AI content | Meta Instagram | Meta Comune | WhatsApp Business | Turtles AI

Highlights

• Meta and UMG extend their agreement to include measures against unauthorized AI content.
• The agreement allows music sharing on WhatsApp and Threads for the first time.
• UMG and RIAA are increasing legal controls against AI misuse in music.
• Meta develops generative AI models with an ethical approach, but concerns about deepfakes arise.

Meta and Universal Music Group strengthen their music-sharing agreement: a new frontier in AI content management

Meta and Universal Music Group (UMG) recently expanded their multi-year collaboration, making it easier and legal for users of platforms like Facebook, Instagram, WhatsApp, Threads, and Horizon to share songs from UMG’s extensive catalog. However, what stands out from this new agreement is the shared commitment to addressing unauthorized AI-generated content.

The primary concerns revolve around the misuse of artists’ works, often employed without consent to train AI models, leading to the creation of fakes that could harm the reputation and rights of creators. Michael Nash, Chief Digital Officer of UMG, stated the company’s commitment to protecting artists’ rights, now and in the future, in collaboration with Meta.

The agreement fits within a broader context of increasing attention to AI technologies by the music industry. In February, UMG removed its music catalog from TikTok due to concerns over online safety and AI misuse. The standoff was resolved only in May, when the two companies reached a compromise that allowed the return of songs by artists like Billie Eilish and Taylor Swift to the platform.

Meanwhile, the U.S. music industry has been the scene of a significant legal battle, with the Recording Industry Association of America (RIAA) suing startups Udio and Suno. These companies allegedly used copyrighted music to train their AI models, a practice that could infringe on the rights of the owners of those works. Sudo admitted to using protected songs but argued that it fell under "fair use."

For its part, Meta has taken an ethical approach to developing generative AI models, such as AudioCraft, MusicGen, and Jasco, claiming they were trained only with Meta-owned or specifically licensed music. However, Meta has also acknowledged that these technologies could potentially be misused to create vocal deepfakes, raising further concerns.

A novel aspect of the agreement is the inclusion of WhatsApp among the platforms where UMG-protected music can be shared, a new feature that could change how users engage with the instant messaging app. Threads, Meta’s social network created to compete with X (formerly Twitter), is also part of the agreement, strengthening UMG’s presence across all Meta platforms.

The relationship between Meta and UMG dates back to 2017 when the partnership first allowed Facebook users to upload and share videos featuring music from the UMG catalog. This new phase of the collaboration seems to aim at a more rigorous and conscious management of AI-generated content, seeking to protect artists in a rapidly evolving digital landscape.

Some questions remain, especially regarding the effectiveness of measures to combat the improper use of music by AI. However, it is clear that the music industry is seeking new ways to protect its rights in the AI era, with strategic partnerships that could become increasingly common.