'Transformative use': Suno's summary-judgment gambit

In March 2026 Suno filed for summary judgment, arguing that training its AI models on copyrighted recordings constitutes transformative use under US copyright law. The motion is a calculated gambit: rather than wait for trial, Suno is asking the court to rule, as a matter of law, that the copying at the heart of its training process falls outside the rights holders' control. Sony's cases against Suno and Udio remain active, with a closely watched hearing expected in mid-2026. Nothing has been decided. That, for anyone with capital or catalogue exposure in this space, is the point.

The AI copyright fair use music debate has moved from academic conference panels to the balance sheets of labels, publishers and the funds that lend against their income. Yet the market's appetite for a headline verdict outpaces the legal timetable. A summary judgment filing is an argument, not an outcome. It signals confidence, or at least a calculated bet on the composition of the bench and the framing of the record. It does not resolve whether wholesale ingestion of commercial recordings for model training sits alongside the kind of transformative use that has previously protected search indexing or parody. That question sits with the court, and the court has not spoken.

Noise versus the underlying asset

For labels and publishers, the temptation is to treat every filing as a data point that moves the value of their catalogues overnight. It does not. What has actually changed since the filing is procedural: a motion has been lodged, and a hearing date sits on the calendar for mid-2026. The recordings themselves, the royalty streams they generate, and the contractual protections already in place around them are unchanged. Conflating litigation news with catalogue value is precisely the kind of noise that a rigorous adviser should filter out before it reaches an investment committee or a lending covenant.

For rights investors and their lenders, the more useful exercise is scenario mapping rather than prediction. What does a licence-based settlement do to future royalty flows from AI training and output? What does an adverse ruling on transformative use do to the negotiating leverage labels currently hold over platforms seeking to license catalogues rather than litigate access to them? Neither scenario is priced with any precision today, and no reputable analysis should claim otherwise. The value of decent legal and financial advice through this period lies not in guessing the outcome, but in building structures, warranties and covenants that hold up whichever way the court leans.

What durable value looks like here

The durable asset in this story is not the litigation outcome itself. It is the underlying architecture of rights: who controls the master, who controls the composition, and how cleanly those rights can be licensed once, if, a framework for AI training emerges. Firms and investors who spend mid-2026 building that clarity, rather than trading on courtroom headlines, will be better placed regardless of how the Suno and Udio hearings resolve.

Courts move on their own schedule. Catalogues, covenants and cash flows should be built to withstand whichever verdict eventually arrives, not to anticipate it.