This week was interesting, watching some of the main collectives and players in the music and technology industries. Let’s see; from YouTube to Ethics on AI, who’s doing the best thinking right now?
#1. YouTube’s latest development on AI and artists’ protection
This is significant for music rights management. That development targets AI-generated content that mimics the likeness or voice of individuals without their permission. After users flag the tracks asked to be removed, YouTube will evaluate takedown requests based on the identifiability of the person, public interest value, and context.
This policy aims to tame the unapproved AI soundalike tracks, complementing similar efforts by major music rightsholders on platforms like Spotify. However, broader legal protections for name, image, and likeness rights remain a work in progress, with ongoing industry and legislative efforts to address these challenges in the best and fairest way possible.
#2. Is there really a slowdown in music streaming subscribers in Sweden?
A decline in paid music streaming subscriptions in Sweden? A few articles lately have left the music industry skeptical at best regarding the numbers of this highly symbolic market for music streaming.
A YouGov survey indicated a drop from 59% to 56% in Premium subscribers from mid-2022 to May 2024. However, internal data from global music companies are found to be contradictory, showing a slight increase in paying subscribers, from 2.80 million at the end of 2022 to 2.87 million in May 2024. The survey’s results are seen as an anomaly, possibly due to its limited sample size.
Industry data also highlights stability in the Swedish market, with Spotify maintaining a significant market share and low churn rates. Despite the survey’s findings, broader evidence, including increased premium on-demand streams and stable subscription revenues, indicates that the Swedish streaming market remains strong and solid.
#3. Clean datasets, at what cost?
The integration of clean datasets for AI music models is becoming crucial as legal battles intensify. As we talked about last week, major record labels are suing AI startups Suno and Udio, accusing them of using copyrighted music without permission.
Without any surprise, this litigation underscores the need for AI companies to use licensed data for training. For instance, YouTube is reportedly negotiating with record labels to license music for its AI models, highlighting a potential path for compliance.
The outcome of these cases will significantly impact the future of AI in music, with the potential for expensive licensing deals to become the norm, ensuring that only well-funded companies can sustain operations. Curious about how the story unfolds? Go deeper on the matter here.
#4. A newcomer in AI ethics
This week, Dublin City University, with funding from the Irish government, launched AI:OK, an initiative to promote responsible AI use in the music industry. AI:OK aims to help identify ethically created music products and services, ensuring AI enhances human creativity while protecting music stakeholders' rights.
The initiative includes self-regulation, transparent standards, and a "trustmark" to label responsibly created music. Partnering with Fairly Trained, which certifies AI models for ethical training, AI:OK co-founder Dr. Martin Clancy emphasized their dedication to ethical AI use, fostering confidence among creators and consumers.
#5. Berlin, from Electronic music to Technology
Let’s wrap up on a lighter note for this week with a deep dive by MusicTech: Berlin's longstanding relationship with electronic music continues to thrive, now embracing AI technology. The city, known for its vibrant techno culture and renowned software startups like Ableton and Native Instruments, clearly has become a hub for AI music experimentation.
Berlin's open, diverse, and artist-friendly environment, coupled with its rich electronic music heritage, truly creates a unique space for AI music development, attracting talent and encouraging groundbreaking projects.
