Last major standing - SONY

January. A Sony Publishing co-write. Two writers, one room, one laptop. One of them typed a working title from the session they had had the week before into Suno. Ninety seconds later it gave them back a passable demo of a song that, by any deal either of them had signed, was not supposed to exist.

They closed the laptop. They kept writing. Nobody called a lawyer, because no songwriter has ever won by calling a lawyer first.

The two settlements that landed last autumn, UMG with Udio in October, WMG with Suno in November, did not include them. The labels signed. The publishers and the writers watched. The rate sheets came back without them on it.

Sony has not signed.

The settlement parade made AI a vendor, not a defendant.

When UMG took the deal, the framing changed. Not "we sued, we won." More: "we found a way to monetise the inevitable." When WMG followed, the inevitability hardened. Suno and Udio became something close to licensees, with all the legitimacy that label implies. The fair-use question, the one the cases were built to answer, slid off the table. No ruling. No precedent. A number on a contract and a press release.

That isn't peace. That's a deferral.

The April filing told us everything.

In April, UMG and Sony filed an objection to the terms of the WMG-Suno settlement. Read that twice. UMG, which had already settled with Udio, joined Sony in challenging what WMG signed with Suno. The cartel didn't hold. The market comp was already a problem, the second it was set.

The mechanism is simple. Settlement one sets the anchor. Settlement two becomes the comp. Settlement three is the cap, and everyone who didn't sign by then is negotiating from underneath their own catalogue. Publishers, songwriters, indies, all of them watching the rate sheet from outside the room.

Sony saw the maths and stayed.

July is the only date that matters.

The summary-judgment hearing in Sony's case against Suno is on the calendar for July. A fair-use ruling is expected through the summer. If the court finds for Sony, the question stops being "what's the rate?" and becomes "do you have any right to be using this catalogue at all?" The asymmetry flips. The settlements look early.

If the court finds for Suno, the comp goes with it. The settlements look prescient. UMG and WMG cashed out at the top of a market that no longer exists.

That is the bet Sony is making. Not on a number. On a ruling

The first deal is a price. The second deal is a precedent.

The publishing catalogue is the part nobody mentions.

Sony Music Publishing is the largest publishing house in the world. Bigger catalogue exposure than either of the majors it competes with on the recorded side. Songs, not masters, are what generative models recombine into output. A publishing-led plaintiff has a different theory of damages, a different theory of harm, a different theory of what licensed even means. Sony's case is built on more catalogue than UMG or WMG had on the table.

Not coincidence. Calculation.

The brave read is the right one.

Settlements are how labels manage AI as a vendor problem. Litigation is how the industry treats AI as a catalogue problem. Sony chose the second framing. The other two majors chose the first. One of those framings ends with a quarterly licensing line item. The other ends with a precedent the next twenty years of model training has to negotiate around.

The risk is real. Bad ruling, and Sony gets the worst of both worlds: a market that has moved on, and case law that makes every plaintiff after them weaker. Good ruling, and Sony resets the entire licensing posture of the industry, and the catalogue owners who weren't in the room get a rate sheet that reflects the work.

Pick a side.

The hearing date is on the calendar. The settlements are on the books. Whichever way the ruling lands, the trade press will be late to it. The version worth reading is the one that says Sony was the only major still asking the question the cases were filed to answer.