Oracle v. Google is over (for now), leaving case law that could change patent damages
After more than two weeks of trial and deliberations, the jury rendered a verdict of noninfringement on all counts in the epic Oracle v. Google litigation. More importantly, on May 29, the district court denied Oracle’s request for a judgment as a matter of law, finding “ample” evidence to support the jury’s decision and dismissal of claims. The next day, the judge also ruled that the Java application programming interfaces (APIs) were not subject to copyright. Taken together, the rulings have led some patent experts to conclude that software—which consists largely of mathematical algorithms—has no value except to provide “a list of ways a lawyer can assert that the patent has been infringed.”
“Patents and software need to get a divorce before somebody gets hurt,” agrees another blogger, who posted daily updates from the courtroom. “The damage from software patents is astounding, and the IP is so puny. There is an imbalance in the legal universe, and it needs fixing. Software is algorithms, and that is mathematics, and it's wrong, totally wrong, to let math be patented. These patents should never have been issued.”
Five Daubert opinions. Oracle has promised to appeal the rulings. In the meantime—and of particular interest to IP appraisers and damages experts—the case has also produced five Daubert opinions on damages: three against the plaintiff’s expert, one against the court-appointed expert, and one against the defendant’s rebuttal experts. Key takeaways include:
• When applying the entire market value of any asserted technology, the expert “must in every case” apportion damages between the patented and unpatented features of the technology; • Current patent law requires only a patent-by-patent apportionment, rather than claim-by-claim (as the court found in one of its prior orders); • Consumer and market share surveys are not “inherently unreliable,” but may become so when the experts (as in this case) “artificially force” the participants or the data to a desired outcome; and • Until they track larger datasets, current patent value studies do not inspire a high level of confidence in their predictive power.