SoundCloud is updating its approach to AI music, with CEO Eliah Seton announcing changes to platform policies aimed at tackling fraud, deceptive activity and copyright concerns. In an open letter, Seton argues that artificial intelligence should support artists rather than replace them. The company is also expanding its creative partnerships with Moises and ElevenLabs while acknowledging a major technical challenge: reliably distinguishing fully AI-generated music from recordings created by musicians using AI-assisted tools.
SoundCloud has outlined a new approach to artificial intelligence, placing artist protection, copyright compliance and greater transparency at the center of its plans. CEO Eliah Seton announced the direction in an open letter addressing both the creative possibilities of AI and the risks emerging as automated music production becomes more accessible.
The announcement includes updates to the platform’s Community Guidelines and Terms of Use, particularly around spam, fraud and misleading activity. It does not introduce a blanket ban on AI-generated music. Instead, the company is attempting to distinguish legitimate creative experimentation from behavior that misleads listeners or violates the rights of musicians.
Seton argues that treating artificial intelligence as entirely beneficial or entirely harmful overlooks the questions artists are asking. For SoundCloud, the challenge is deciding which uses deserve support, how rights should be protected and what information listeners need when they encounter music made with automated tools.
The company’s existing public policies already restrict unauthorized impersonation, including using AI to imitate someone’s voice or likeness. They also address copyright infringement and attempts to manipulate accounts or engagement. The letter presents a broader commitment to enforcement without providing a complete, itemized comparison of every amended provision.
One important limitation concerns timing. The publicly available Terms of Use identify August 17, 2026, as their latest amendment date. That means the exact wording and effective dates of any additional changes announced in November require further confirmation from the company.
The issue of artist consent has particular significance for SoundCloud following controversy over its AI-related terms in 2025. Musicians raised concerns that uploaded recordings might be used to train generative systems without sufficient permission, prompting the company to clarify its position.
In May 2025, SoundCloud committed not to use creators’ content to train generative AI models designed to reproduce their music, voice or likeness without explicit consent. That protection is therefore part of the platform’s established policy history, rather than a provision being introduced for the first time in this announcement.
The company is simultaneously increasing its involvement with AI-powered music tools. Its partnership with Moises was announced on October 21, 2026, followed by a collaboration with ElevenMusic from ElevenLabs on November 5. Both agreements are being presented as part of the wider strategy rather than newly unveiled partnerships.
Moises offers tools that musicians can use during audio production and creative development, while ElevenMusic provides generative music capabilities. SoundCloud sees such services as potential connections between the stages of making music, collaborating with others and publishing the finished result to an audience.
Seton says the platform now works with more than ten approved AI partners. According to the company’s stated criteria, these partners are expected to respect creator rights, maintain transparency and follow responsible approaches to training technology. Those standards remain company commitments, not independent verification of every service’s practices.
Perhaps the most revealing part of the letter concerns the difficulty of identifying AI music. SoundCloud acknowledges that there is no simple technical boundary between a track created entirely by a generative system and one made by a human using AI tools during production.
A musician might use automated software for editing, separating sounds or developing an arrangement while remaining responsible for the composition and performance. A different recording could be generated almost entirely from a text instruction. Both involve artificial intelligence, but the creative processes behind them are substantially different.
This distinction complicates labeling. SoundCloud says its approved partner integrations can display identification information supplied by those services, helping listeners understand where a recording originated. However, that approach depends partly on the information provided by the partners and does not amount to universal detection.
For content arriving through other channels, the platform uses technology, internal review processes and its existing policies to assess possible problems. Seton acknowledges that industry standards for detecting and labeling AI-generated music are still developing, leaving unresolved questions about accuracy and consistency.
The strategy also extends to music discovery. SoundCloud has used machine learning for personalized recommendations for years, including capabilities developed through its acquisition of Musiio. The company wants to improve how listeners find emerging artists while retaining the sense of discovery associated with the service.
Seton also describes greater user control as a future priority. However, the announcement does not confirm the immediate launch of a universal filter allowing fans to remove every AI-generated recording from their recommendations. More personalized choices remain a direction for further development.
SoundCloud says its approach is informed by conversations with creators, including artist roundtables and exchanges across communities such as Reddit and Discord. The intention, according to Seton, is to incorporate those perspectives into product development, partnerships and decisions about platform rules.
The practical questions now concern enforcement and accountability. Artists need ways to challenge unauthorized imitation and copyright violations, while listeners need accurate information about music’s origins. The company has outlined its principles, but the effectiveness of its approach will depend on how clearly those protections are implemented.
For SoundCloud, the next phase of AI music is therefore about more than offering new production tools. It involves establishing whether artist consent, transparent identification and trustworthy discovery can keep pace with the growing volume of music created using artificial intelligence.



