The Twisted History of Oracle v. Google

Oracle v. Google began in 2010. The Supreme Court issued its decision on fair use on April 5, 2021, eleven years later. In between there was a lot of discovery, and multiple trial proceedings: one in which a jury rejected Oracle’s patent claims, a second in which the court reserved for itself copyrightability issues but the jury decided infringement and fair use being instructed to assume the APIs were copyrightable. The jury found infringement but deadlocked on fair use. The trial judge (Alsup, J.) then ruled the APIs weren’t copyrightable, thereby rendering the jury’s verdicts moot.

There was an appeal to the Federal Circuit (due there being a patent claim even though the patent claim wasn’t appealed). The CAFC reversed on copyrightability but remanded for a new jury on fair use, saying the jury needed to find facts. A petition for certiorari to the Supreme Court was denied. 

There was then a second trial on fair use which lasted a week. The jury deliberated for three days and gave a general verdict of fair use.  There was a second appeal to the Federal Circuit which, inexplicably in light of having remanded to a jury for factual findings, vacated the jury’s verdict and held for Oracle even though the jury found no facts. Nothing had changed since the CAFC’s remand due to an alleged need for jury fact findings. With a general verdict there is just “yes” or “no.” There was then a successful petition for certiorari over the Solicitor General’s advice not to take the case. The April 5, 2021 opinion for the eight member Court was by Justice Breyer, over a dissent by Justices Thomas and Alito. The Court had heard two issues, copyrightability and fair use, but only decided fair use because they were split 4-4 on copyrightability. 

This blog discusses my thoughts at the time as copyright counsel to Google for the strategy to prevail at SCOTUS. I don’t reveal any secrets or privileged information or any other person’s thoughts. The goal is merely to show how serendipitous the outcome may have been.

When the petition for cert. was granted, Justice Alito owned stock in Oracle. He had recused himself from other cases involving Oracle for that reason and I assumed he would do so again. That would have led to an 8 member Court, and if there was a 4-4 split the CAFC’s decision adverse to Google would have stood. The Court had split twice before on fair use, in the Jack Benny case and in the Williams & Wilkins photocopying case. The jury in the first trial was also deadlocked on fair use. The Supreme Court had split also split in 4-4 Lotus v. Borland, a case which loomed large in Oracle. It loomed large for a few reasons: first, it dealt with issues that were similar to those in Oracle. Second, it came from the First Circuit, where Justice Breyer had sat on before becoming a Supreme Court Justice and it was not a secret that he agreed with Judge Boudin’s concurrence in that case. Moreover, then Professor Breyer had in 1970 written his tenure piece at Harvard Law School on related issues, entitled “The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs.”  84 Harv. L. Rev. 281. Justice Breyer had been waiting 51 years for his chance to write the Big Opinion on copyright and computer programs. Oracle was the perfect vehicle for that, and of course, an influential vote for Google.

Justice Ginsburg, though ill, was still on the Court when the Court granted cert. and when the briefs were filed. I had known her almost 35 years at that point, and she had cited me in most of her copyright opinions. It was her vote that I believed to be a critical one. I believed she would find the APIs copyrightable, and thus she only needed three more votes for a ruling that would leave in place the favorable to Oracle CAFC opinion. Google, by contrast , needed five votes to come out ahead. Oracle in this scenario had an automatic advantage in votes. I had assumed Alito would be a vote for Google given his opinion for the Third Circuit in the Southco parts case, but I also assumed he would recuse due to owning Oracle stock. I was doubtful that RBG would de novo agree the use was a fair use, but at the same time she was very respectful of juries as well as civil procedure. 

Since this was a general verdict I wanted to couch the brief on fair use that way: this was a jury verdict, reached after a one week trial and three days of deliberations. I also hoped this would persuade others to join, especially Justice Thomas who in his opinion for the Court in Feltner v. Columbia Pictures Television, 523 U.S. 340 (1998) had held there was a Seventh Amendment right to a jury for statutory damages, even though title 17 referred to the “court” awarding them. In addition to the Seventh Amendment, there was what I thought was a surefire civil procedure based argument for upholding the jury verdict: Federal Rule of Civil Procedure 39(c)(2), which provides that even where an issue is not triable to a jury as of right, the district court “may, with the parties' consent, try any issue by a jury whose verdict has the same effect as if a jury trial had been a matter of right.” Oracle was tried to a jury with the parties' consent, both times. FRCP 39(c)(2) required that the jury's verdict have the same effect as if it was of right. This provision was cited to the Supreme Court. 

Had RBG written the majority opinion, I thought she would have agreed with the CAFC on copyrightability (infringement hadn’t been appealed). She might express some skepticism about fair use on the merits, but say that the Court was not taking a position on whether it was fair use, and was letting the jury verdict stand because there were sufficient grounds for the general verdict, which under either the Seventh Amendment or FRCP 39[c](2) had to be respected. 

What happened was quite different. Justice Alito sold his Oracle stock. RBG died on September 18, 2020, 19 days before the Oracle oral argument on October 17. leaving an 8 person court. (Justice Barrett took office on October 27, and did not participate in the case). The court could not reach a majority on copyrightability, splitting 4-4 and but did reach a majority on fair use, 6-2, Justices Thomas and Alito dissenting. My strategizing on fair use was fine but not on the jury part of it, ostensibly, since Justice Breyer pooh-poohed the Seventh Amendment argument, and ignored FRCP 39[c][2], although he repeatedly emphasized the evidence before the jury and what the jury could have decided.. Justices Thomas and Alito would have reached copyrightability and found the works copyrightable, and the use not a fair use.

Had RBG not died I think it likely the Court would have voted at least 5-4 in favor of copyrightability, and RBG would have written the opinion, with at least Justice Breyer in dissent on that point. On fair use, we know there were at least two against it, Thomas and Alito. Perhaps RBG would have been as well which would have made three but that would have led to complications in writing the majority opinion, so perhaps someone beside RBG would have written the majority opinion, but it wouldn't have been Breyer since he would have been on the losing side of copyrightability. Only one other justice would have been needed to make it a 4-4 split on fair use. The actual vote of 6-2 on fair use might show it would have been 6-3 even with RBG’s vote against fair use, but who knows?

.