A reluctance to apply strict examination requirements to software opened up the floodgates to questionable software patents in the late 1990s. The requirements of the U.S. patent code that police scope and notice were not considered significant hurdles in software. Perhaps most significantly, the Section 112 requirements of written description and__ definiteness__ were not applied vigorously to software.
But borrowing from the biopharmaceutical industry, and particularly the bioinformatics context, I’d argue that both written description and definiteness could readily be used – right now, without any change in law – to tame claims in software.
It’s actually rather ironic that the biopharmaceutical industry, traditionally seen as different from software, can now offer the software industry some lessons.
In biopharmaceuticals, an application fails the written description requirement if its disclosure provides insufficient physical structure for its claims. So inventors can’t typically claim solely by function; for example, they can’t claim a method for chemically inducing a particular effect unless they reveal at least one or more actual chemical formulae. As for definiteness, an application fails that requirement if its claim language doesn’t clearly delineate the inventive territory at issue.
Lessons From Bioinformatics: Written Description and Definiteness
Now consider bioinformatics: It’s just software (computer science, algorithms) applied to the biopharmaceutical industry. Here too, both written description and definiteness have long been used by the sub-unit within the Patent and Trademark Office (PTO Art Unit 1631) that examines bioinformatics applications to narrow scope and improve notice.
#### Arti Rai
##### About
Arti Rai is a professor at Duke Law School and the Duke Institute for Genome Sciences & Policy. She has also taught at the Harvard, Yale, and University of Pennsylvania law schools. Rai served as Administrator of the Office of External Affairs at USPTO (2009-2010), where she led policy analysis of the patent reform legislation that ultimately became the America Invents Act. Rai won the World Technology Network Award for Law in 2011.
Written description is used to reject applications that claim broad function without providing disclosure of a specific associated algorithm: for example, a 2003 patent application (#10/620,796) in Art Unit 1631 claiming a bioinformatics tool for finding “translation initiation codons” within DNA sequence data. Translation initiation codons are quite important because they signal the beginning of the actual protein-coding region within a DNA sequence. So though the idea of using algorithmic analysis to find such codons is appealing, the examiner rejected this patent for failing to describe the details of the actual algorithm used. And that’s as it should be.
Art Unit 1631 has also routinely used the definiteness requirement to reject applications for using terms such as “mapping data” (#10/332,999) that are unclear even in context. The examiners there recognized early on that they didn’t need to follow the very lax “insolubly ambiguous” standard articulated by the Federal Circuit (in the 2001 case of Exxon Engineering) for cases that came before courts.
Unlike many other patentability requirements, both written description and definiteness are relatively easy-to-implement mechanisms for limiting broad claiming and improving notice. They provide overworked and under-resourced patent examiners the right tools to do their job. Although the Federal Circuit’s initial formulation of written description in University of California v. Eli Lilly (1997) was overly rigid, subsequent refinements by the court and by the PTO yielded a doctrine that appropriately tames claims.
I’m backing this up with more than just anecdotal evidence. In a study (conducted with Bhaven Sampat) of patent applications filed in 2003, we found:
- 77% of the applications in Art Unit 1631had Section 112 rejections. Many of the Section 112 rejections invoked written description in a manner similar to the example above.
- In contrast, almost none of the applications in the control software Art Unit (Art Unit 2123) had written description rejections.
- Art Unit 1631 deployed the definiteness requirement criterion in a truly remarkable 95% of its Section 112 rejections. This compares with roughly 50% for the control software Art Unit.
PTO Art Unit 1631 therefore provides many worthwhile examples of how other examiners could readily utilize written description and definiteness to tame software claiming. The failure of other software-examining Art Units to vigorously adopt written description and definiteness is an unfortunate example of institutional learning delay by the U.S. Patent Trademark Office.
But in recent years the USPTO has become a passionate advocate of both written description and definiteness across the board. In fact, it filed a key amicus brief in Ariad v. Eli Lilly, a 2010 case in which the Federal Circuit made clear that written description is required for all technologies, because:
Moreover, written description and definiteness aren’t good just for patent application examiners. They represent administrable tools for challenging and judging the validity of already-issued software patents. By law, the newly created Patent Trial and Appeals Board (PTAB) must conclude proceedings within a year; written description and definiteness could provide challengers and PTAB judges with just the tools they need.
Editor's Note: Given the enormous influence of patents on technology and business -- and complexity of the issues involved -- Wired is running a special series of expert opinions on "the patent fix". To help move reform efforts forward, s*ome of these proposals also advocate specific Solutions to the Software Patent Problem (as part of a conference hosted by the Santa Clara University High Tech Law Institute). *
Editor: Sonal Chokshi @smc90
