SoundCloud Revises Policies to Allow AI Training with User Content
SoundCloud appears to have quietly revised its terms of service, now allowing the company to use user-uploaded audio for AI training purposes.
Tech ethicist Ed Newton-Rex pointed out that the updated terms give SoundCloud the authority to utilize uploaded content for “informing, training, [or] developing” AI.
“You explicitly consent that your Content may be used to inform, train, develop, or serve as input to artificial intelligence or machine intelligence technologies or services as part of and for providing the services,” the terms stated, with the update dated February 7.
The terms also feature exceptions for content covered by “separate agreements” with third-party rights holders, such as record labels. SoundCloud holds various licensing agreements with numerous independent labels and major music publishers like Universal Music and Warner Music Group.
TechCrunch could not find a clear opt-out option in the platform’s online settings menu. SoundCloud has yet to respond to a request for comment.
Like many major creator platforms, SoundCloud is increasingly adopting artificial intelligence.
Last year, SoundCloud partnered with almost a dozen vendors to roll out AI-driven tools for remixing, generating vocals, and producing custom samples on its platform. In a blog post last fall, SoundCloud mentioned that these partners would access content ID solutions to “ensure rights holders receive proper credit and compensation,” committing to “upholding ethical and transparent AI practices that respect creators’ rights.”
TechCrunch event
Berkeley, CA
|
June 5
BOOK NOW
Recently, several content-hosting and social media platforms have changed their policies to include provisions for both first- and third-party AI training. In October, Elon Musk’s X updated its privacy policy to allow external companies to train AI using user posts. Last September, LinkedIn modified its terms to allow scraping of user data for AI training purposes. Additionally, in December, YouTube started enabling third parties to train AI with user clips.
These adjustments have ignited considerable backlash from users who argue that AI training policies should be opt-in rather than opt-out and assert that they have the right to be credited and compensated for their contributions to AI training datasets.
Updated 2:22 p.m. Pacific: A representative from SoundCloud provided an email statement, parts of which are included below:
“SoundCloud has never used artist content to train AI models, nor do we develop AI tools or allow third parties to scrape or use SoundCloud content from our platform for AI training purposes. We have put technical safeguards in place, including a ‘no AI’ tag on our site to explicitly prohibit unauthorized use.
The February 2024 update to our terms of service aimed to clarify interactions between content and AI technologies within SoundCloud’s platform. Use cases include personalized recommendations, content organization, fraud detection, and enhancements to content identification with AI assistance.
Any future AI applications at SoundCloud will be crafted to support human artists by improving their tools, capabilities, reach, and opportunities on our platform. Examples include refining music recommendations, generating playlists, organizing content, and detecting fraudulent activities. These initiatives align with current licensing agreements and ethical standards. Tools offered by partners like Musiio are strictly for artist discovery and content organization, not for training generative AI models.
We acknowledge the concerns raised and remain dedicated to open communication. Artists will continue to maintain control over their work, and we will keep our community updated as we responsibly investigate innovation and apply AI technologies, especially as legal and commercial standards evolve.”
