Software Patent Eligibility
posted by David Schwartz
This post reflects my initial impressions of an important Federal Circuit development in patent law, which is my primary area of scholarly focus. On Friday, the Federal Circuit, sitting en banc, ruled on a controversial and divisive patent law issue, whether software inventions are patent eligible subject matter. Unfortunately, I find the decision in this case, CLS Bank v. Alice Corp., quite unsatisfying.
The court, sitting with 10 judges, issued 7 separate opinions spanning 135 pages. The court only agreed upon a very brief – 55 words – per curiam opinion affirming the district court ruling that the asserted patents were invalid. The per curiam opinion explained that the “method” and “computer readable media” claims were deemed not patent eligible by the Federal Circuit, while the court was equally divided on the status of the “system” claims. (Basically, there are several different ways that a software invention can be claimed in a patent, including as a process/method of performing steps; as software embedded upon a computer readable medium (i.e., a DVD); and as a system (i.e., software running on a machine/computer).) None of the remaining substantive opinions garnered more than 5 votes – thus, none are binding precedent. Although a majority of the Federal Circuit judges found the method and media claims invalid, a majority could not agree upon the reasoning. Below I will briefly provide a few preliminary observations about the opinions.
To provide some context, in the 1970s and 1980s, it was unclear whether software was patent eligible. However, by the mid- to late-1990s, the U.S. Patent Office and the Federal Circuit both decided that software was patent eligible, so long as it was properly claimed and it satisfied the other requirements of patentability (i.e., the invention was novel, non-obvious, etc.) More recently, the Supreme Court and many in the software community pushed back. In several cases not involving software, the Supreme Court disagreed (here, here, and here) with the Federal Circuit’s permissive standards regarding patent eligible subject matter, causing a gradual reconsideration of the doctrine. Opponents of software patents publicly argued that software was different from other inventions for several reasons, including because product life cycles were shorter and the claimed inventions would have been developed without the incentive of a patent.
After a series of split Federal Circuit panel decisions, the entire Federal Circuit agreed to consider the issue of patentability of software en banc in late 2012. The CLS Bank case was closely watched by the software industry and practicing bar. Last Friday’s decision contains 6 written opinions totaling over 130 pages, all of which are important but technically dicta. These include a concurring opinion (joined by 5 judges); a concurrence-in-part, dissent-in-part (joined by 4 judges); a dissent-in-part (joined by 4 judges); a concurrence-in-part (1 judge); a dissent (2 judges); and “additional reflection” (1 judge).
The opinions are long and will require some time to digest and study. However, I offer the following initial observations:
1. The en banc ruling came fast by historical standards. It took slightly over 3 months from oral argument to written opinion. This is faster than the typical precedential panel opinion, let alone one with 6 separate written opinions on an important issue. It is somewhat surprising that the court issued such a splintered and long opinion in a short time period. Since a majority of judges agreed that certain patent claims were patent ineligible, it would be have preferable if they agreed upon an opinion that explained a common rationale. Without a majority opinion, the future impact of the majority’s ruling is less clear.
2. The Federal Circuit was originally created in part to create uniformity in patent law. Unfortunately, the Federal Circuit itself seems internally conflicted on the limits of patent-eligible subject matter, including of software patents. As the CLS Bank case illustrates, a majority of the Federal Circuit can’t agree on how to approach software. Absent Supreme Court intervention, whether a particular software invention will be knocked out or upheld apparently will depend upon the panel assignment on appeal. This is a particularly bad sort of uncertainty since it requires litigation through appeal for the result to be predictable.
3. The divisions in the court, along with particularly sharply worded dissents and the importance of software patents more generally, may peak the Supreme Court’s interest. For example, in her dissent, Judge Moore stated that if the 5-judge concurring opinion were adopted as the law, then every patent covering “inventions directed to computer software or to hardware that implements software” would be not patent eligible. According to Judge Moore’s math, that includes over 300,000 patents in total and nearly 20% of the total number of patents issued in 2011. She argued that the opinion “would decimate the electronics and software industry… There has never been a case which could do more damage to the patent system than this one.”
4. It is also somewhat surprising that the Federal Circuit decided this case without waiting for the Supreme Court’s pending decision in another case involving the issue of patent eligibility. That case (Association for Molecular Pathology v. Myriad Genetics) does not deal with computer software, but rather inventions related to human genes. It is quite possible that the Supreme Court may GVR (grant, vacate, and remand) CLS Bank after a cert petition is filed in light of its decision in Myriad.
5. The Patent Office follows the Federal Circuit’s opinions for guidance on patent examination. Given that lawyers will have difficulty determining the state of the law after CLS Bank, the Patent Office also likely will struggle to provide useful examination guidelines to patent examiners on software patents.
6. The Federal Circuit’s membership is rapidly changing. The newest Federal Circuit judge, Judge Taranto, was not confirmed until after the oral argument. He did not participate in the case. Furthermore, there are two pending nominees to the Federal Circuit, Raymond Chen and Todd Hughes. Assuming Chen and Hughes are confirmed at some point, these three judges may hold the balance of power on software patentability. Furthermore, Senior Judge Richard Linn participated on the CLS Bank en banc panel because he was a member of the original 3-judge panel for the case. He was one of the 5 judges believing that the system claims were patent eligible. As a senior judge, it is unlikely that he will be on any future en banc decisions regarding patent eligibility.
7. The underlying ruling is that the software claims at issue are not patent eligible. This case runs against the view of most academics (myself excluded) that the Federal Circuit is reflexively pro-patentee.