Suno answered the RIAA suit by comparing training to learning

The system

When major record labels represented by the RIAA sued Suno on 24 June 2024, alleging its training data infringed their recordings, Suno answered in public rather than only in court filings. Co-founder and CEO Mikey Shulman published The Future of Music on Suno's own blog on 1 August 2024, addressing the lawsuit directly as 'an update on the recent legal action from the RIAA and what is at stake.'

What the documents establish

Shulman's post states Suno's position without using the phrase 'fair use' but arguing the underlying theory: that Suno trains on 'medium- and high-quality music we can find on the open internet,' the same approach the post says other AI companies use, and that this training process resembles 'a kid learning to write new rock songs by listening religiously to rock music' rather than copying. The post calls the labels' central claim, that the models 'parrot' copyrighted material, a misconception, and states that Suno has 'never allowed prompts for specific artists or copyrighted songs.' A later Suno post, How We're Building the Future of Music Responsibly, dated 6 August 2026, restates the same underlying claim in more concrete terms, describing a strategy the company calls 'Original Creation, By Design,' including the decision not to use artist names in training metadata and partnerships with Audible Magic and Musixmatch to screen uploads for unauthorized material. Neither post is a court filing; both are Suno's own public description of its position, published on its own platform.

Craft and rights

For a producer weighing whether to use Suno commercially, the company's own account draws a line between training on 'the open internet,' which it defends as learning, and outputs that reproduce a specific existing song or artist, which it says its guardrails are built to prevent. What remains a legal question, not a settled fact, is whether a court agrees that the training analogy holds under the Copyright Act's fair-use factors; Suno's blog post is the company's litigation position and public argument, not a judicial finding, and this report treats it as exactly that.

Outcomes and open questions

The RIAA's suit against Suno was pending at the time of Shulman's response, and its resolution is not established by these blog posts. Whether 'learning is not infringing,' as the post puts it, holds up as a matter of law depends on rulings this report has not verified as final.

• Does Suno's account of blocking artist-name prompts match your own experience generating with the tool?
• Would a court accepting the learning analogy change how you evaluate other AI music tools' training claims?
• What does the newer watermarking and transparency roadmap add that the original 2024 response did not promise?

A company's defense of its own training practice, however detailed, is an argument for its side of a live dispute, and reading it as a verdict gets the genre of the document wrong.