SoundCloud Updates Terms of Service Following Backlash to AI Policy

by

SoundCloud recently revised its Terms of Service following criticism from musicians and activists about its policy regarding AI training in music.

In a public letter issued on Wednesday, SoundCloud CEO Eliah Seton declared that SoundCloud has never employed artist content to develop AI models, a claim which mirrors what a SoundCloud representative conveyed to The Hollywood Reporter last week.

Seton clarified that our AI tools are not designed for creating music, large language models, or imitating human work. They’re not meant to replace you. It’s a firm ‘no’, period. Furthermore, we don’t develop generative AI and we don’t let third parties use SoundCloud content from artists to train these AI tools.

The opposition arises due to a change in SoundCloud’s terms of service, effective from February 2024. This update allows the company to use users’ content to educate, train, improve, or be used as data for artificial intelligence (AI) or machine learning technologies or services, with the aim of enhancing their provided services.

Last week, an update sparked discussions and criticism from musicians concerned about their work being employed in training artificial intelligence models for generative purposes. In a statement reminiscent of SoundCloud’s last week, Seton mentioned in the letter that the platform revised its terms of service to explain how they might use AI internally to enhance the user experience, specifically by improving search functionality, playlist creation, and content suggestions, all aimed at benefiting both artists and fans.

Seton expressed on Wednesday that the wording in the update was “overly general and lacked sufficient clarity.

“It created confusion, and that’s on us,” Seton wrote. “That’s why we’re fixing it.”

According to the new update in SoundCloud’s terms of service, they will not employ your content to develop AI models that mimic or generate your voice, music, or likeness without your explicit approval. This approval must be given voluntarily through an opt-in process. The previous wording will no longer apply.

Seton proposes that, if there’s a chance to employ generative AI to support and elevate our human artists, we could offer this possibility, only after obtaining their clear agreement. This would be done through an ‘opt-in’ process where the artists themselves choose to participate.

Seton’s stance is clear: AI should complement artists rather than supplant them. The way we apply these tools on SoundCloud will consistently prioritize this idea. AI is set to transform the music industry, offering fresh possibilities but also significant hurdles. To address these issues effectively, our strategy will always revolve around one core value: prioritizing the artist.

SoundCloud’s latest decision arises amidst ongoing debates surrounding Artificial Intelligence within the realms of music and entertainment, a topic that has sparked controversy, as evidenced by the recent dismissal of the Copyright Office’s Director, Shira Perlmutter, over the weekend.

As a gamer, I’ve noticed the recent changes, but they don’t seem to have completely addressed everyone’s concerns. Ed Newton-Rex, founder of Fairly Trained, a music advocacy group, who stated last week he would be pulling his music from this platform, tweeted on Wednesday that the update doesn’t go far enough for him.

In simpler terms, Newton-Rex stated that the new policies mention they won’t create AI models mimicking a person’s voice or style directly, but they may still develop AI systems based on someone’s work, which could potentially compete with them in the market. To truly alleviate concerns about using personal content, the policies should explicitly state, “We will not use Your Content to train generative AI models without your explicit permission.

Read Seton’s full letter here.

Read More

2025-05-15 04:24