Peer to Patent
Community Patent Review
Project Manager, Peer-to-Patent: Community Patent Review; Student Research Fellow, Institute for Information Law & Policy, New York Law School.1
When examiners review a patent application, they compare the invention to vast amounts of publicly available information in order to decide whether the new invention is enough of an innovation to deserve being granted a patent and all the perks that come along with it. Patent examiners, under extreme workload pressure, are granting overly broad and non-meritorious patents. However, the problem is not the result of any actual ineptitude on the part of the examiners. Rather, it is generally a consequence of unfocused information, where examiners are either forced to sift through too much information to find something relevant or are unable to find something relevant because there is not enough information to sift through.
The United States Patent and Trademark Office (USPTO), the Federal Trade Commission, and the National Research Council have each conducted their own investigations into the perceived problems underlying the current state of patents.2 While adopting different approaches, all three of the investigations agree that something must be done to improve the quality of both patent examination and the information available to patent examiners. As part of the USPTO’s own commitment to improving the patent system, it recently launched a year-long pilot of “Peer-to-Patent: Community Patent Review” in collaboration with New York Law School. Peer-to-Patent is a program designed to address directly the information deficit. The program actively seeks out information that may not reside within the closed USPTO search system. The goal is to bring transparency to an otherwise opaque process and the right information to those making crucial decisions.
On April 18, 2007, the Patent Reform Act of 2007 (PRA) was introduced to the 110th US Congress. Less than two weeks later, on April 30,3 the US Supreme Court handed down a major, unanimous decision in KSR International Co. v Teleflex Inc. (KSR) regarding the non-obviousness element of patentability. The former has been billed as “the most significant patent reform legislation in over 50 years”, while the latter ostensibly does away with a standard developed by the Court of Appeals for the Federal Circuit that has been used for nearly two decades of patent litigation and has been relied upon by the USPTO in the prosecution of over 2 million patents. Both are very significant departures from our traditional system of patenting and mark a new trend towards confronting problems inherent in our patent system.
But these reform efforts, despite their far-reaching sweep, are unlikely to move forward with any alacrity. There are a number of reasons for this. From a philosophical perspective, the institution of patents is explicitly mandated by the US Constitution,4 making it part of the “supreme law of the land”,5 any changes to which the US government has traditionally approached with trepidation. Further apprehension can be attributed to more practical factors. Consequences associated with economic tinkering, disagreement between industry and special interest groups regarding reform measures, and lack of public interest in the topic are just a few of the obstacles standing in the way of a serious overhaul. Thus, while both the PRA and KSR decision signal the possibility of major change for the current patent system, that change is not likely to occur in the near future.
Versions of the PRA recently moved through both the House and Senate Judiciary Committees relatively unscathed but opponents of the act remain skeptical and passage by the entire Congress is far from certain. Noted patent commentator Dennis Crouch suggested on his blog, Patently-O, that, as a result of its “top secret development”, the PRA is “so one-sided that it is quite unlikely to move forward without a complete overhaul.”6 The Innovation Alliance7 issued a statement expressing “[disappointment] with the lack of real progress made by the House Judiciary Committee on addressing key issues of concern,”8 and another saying that “as passed by the Senate Judiciary Committee, the Patent Reform Act is still very problematic”.9 While there are just as many who are convinced that the PRA (in some form) will be passed (and soon)10, the stark contrast between pro and opposing sentiment illustrates the lack of consensus with regard to the legislative solution and points to just how far away real compromise and progress may be.11 History seems to also teach against placing too many eggs in the PRA basket. Comparable omnibus patent reform legislation has been proposed each of the last two years,12 none of which has gone anywhere before dying out in Congress.
Similarly, the decision handed down in KSR v Teleflex, though unanimous among the Supreme Court justices, was far more restrained in its wording than many thought it would or should be. During oral arguments in November of 2006, Justice Scalia referred to the Court of Appeals’ teaching-suggestion-motivation (TSM) test for nonobviousness as “gobbledygook” and “irrational”13. But, in the actual opinion delivered 5 months later, Justice Kennedy, writing for the Court, was much kinder, and the Court, seemingly, took issue more with the application of the test than the test itself.14 While the overall opinion was still somewhat pointed, this was a far cry from Chief Justice Roberts’ criticism of the TSM test as “worse than meaningless”.15 It remains to be seen how the lower courts, such as the Federal Circuit, will come to apply the Supreme Court’s holding.
The point in all this is that while parties on all sides of the patent debate anticipate changes, there is a reluctance to rock the patent boat too much. And though the PRA and the KSR decision both provide hope that we are moving in the right direction toward reform, it’s clear that we need to embrace intermediate measures as a means to that improved end.
The Patent Examination Process
The two high-profile events examined above seem to signify a renewed willingness by the US government to tackle controversial, modern patent issues (or at least an inability to avoid tackling those issues any longer). In either case, it’s become evident that many of the problems inherent in our system of patents are beginning to be addressed. However, while both the PRA legislation and KSR decision attempt to reconcile perceived ills in the current law, neither resolves what is possibly the most fundamental aspect of the overall patent problem: the patent examination process.
The main concern is that the demands placed upon the United States Patent and Trademark Office (USPTO) have resulted in granting overly broad and non-meritorious patents. Today 5,500 patent examiners labor independently, under a backlog approaching 1 million applications, with no more than 18-20 hours to review each one.16 The number of patent applications filed per year has grown steadily from 250,000 in 2000 to over 400,000 in 2006.17 If no action is taken, the backlog is projected to reach 1.4 million applications by 2012.18 These numbers are in stark contrast to the European Patent Office (EPO) whose 3,500 examiners received 208,000 patent applications in 2006 while working under a backlog only about 1/3 that of the USPTO.19 Though informative of the problem, these numbers are not conclusive. The third of the “Trilateral Offices”, the Japanese Patent Office (JPO), works under similar, if not more, pressure than the USPTO, having received 400,000+ patent applications annually since 1998 while maintaining a backlog of about 750,000. However, the JPO employs roughly 1/3 as many patent examiners as the USPTO (1,358).20
The problem is likely a combination of understaffing and lack of information. The former is seemingly easy to rectify: hire more examiners. The USPTO is doing exactly that. In 2006, the USPTO hired 1,19321 new patent examiners and plans to hire an additional 1,200 examiners this year and each year for the next five years.22 Lack of information is a more important problem and will be far more difficult to solve.
Getting the right information to patent examiners is a critical component of the examination process. Ideally, when a patent application is filed, an examiner compares the stated invention against everything that has come before it: what is called “prior art”. The prior art is then used to decide whether an invention is novel and non-obvious (among other things) and therefore worthy of a patent. The initial burden of showing that a claimed invention is obvious, and thus not patentable, is statutorily placed upon the examiner.23 To balance that burden, the Code of Federal Regulations places on applicants “a duty of candor and good faith in dealing with the Office, which includes a duty to disclose to the [USPTO] all information known to that individual to be material to patentability”.24 However, even acting in good faith, a patent applicant cannot disclose what he or she is unaware of and thus any disclosures should not be treated as exhaustive. Yet, due to the aforementioned time constraints, they often are. Case in point, the patent that the Supreme Court eventually relied upon in reaching its decision in KSR was not disclosed as prior art in the patent application of the patent at issue.25
The resources available to the patent examiner are finite, partially due to security precautions. The examiners options are limited to internal sources available at the office, such as the three USPTO computer systems (EAST, WEST, and FPAS) and other proprietary database libraries.26 This reliance on such a limited pool of knowledge presupposes a well-documented history of innovation and has the practical effect of placing the examiner at a disadvantage when searching for relevant prior art. In many cases the pile of hay is too large to find a needle and in other cases, the hay is not in a pile to begin with.27
The Effects of the Patent Reform Act and the KSR Decision
To complicate matters further, the KSR decision has the potential to increase the workload placed upon patent examiners. Prior art as a concept expansively includes everything known to man up to the point of any given invention. Quite obviously, however, not all prior art will actually be relevant. The purpose of the Court of Appeals’ TSM test, the Graham test supported by the Supreme Court, and a myriad of other analyses developed by courts over the years, is essentially to determine what prior art is relevant and how it should be considered. While the USPTO is not controlled by any particular court decision, the tests used by courts no doubt give the USPTO guidance as to how they should be examining patent applications. To wit, the USPTO has already sent the Office of Management and Budget a draft of training guidelines for use by patent examiners in light of the decision handed down in KSR.28
The standard for determining non-obviousness is a “person having ordinary skill in the art”29 (PHOSITA). That is to say that if the new invention departs from the prior art in an obvious manner, the invention is not patentable.30 In the Court of Appeals’ view, the PHOSITA “attempting to solve a problem would likely be led only to those elements of prior art designed to solve the same problem”.31 However, under the Supreme Court’s analysis in KSR, “a person of ordinary skill is also a person of ordinary creativity, not an automaton”.32 A logical reading of this appears to widen the scope of relevant prior art to include items that one might not normally associate with the field of the claimed invention, thus adding to the existing patent examiner workload.
Additionally, the PRA proposes changing patents from a first-to-invent system to first-to-file. In doing so, the PRA would amend sections of the US patent law33 so that examiners must include public use and on sale activity outside the US into their calculations the same way they would consider those types of prior art from domestic sources.34 This would further widen the scope of prior art that may have bearing on the ultimate outcome of a claimed invention’s patentability.
To be clear, expanding the pool of potentially relevant prior art is not a bad thing. In fact it is sensible to compare claimed inventions against as much prior art as possible. But increasing the breadth of prior art, ceteris paribus, diminishes the ability of an examiner to “get it right” because it broadens the limits of what counts as prior art. Examiners can only digest and appreciate so much prior art in their allotted time, regardless of how much prior art they //should //be looking at. Adding examiners will help, but those examiners need to have access to the right information or else any attempt at a solution becomes an exercise in futility. The USPTO wants to reduce patent pendency time while increasing the quality of patents issued. This is a feat that can only be accomplished by addressing the information deficit.
Peer-to-Patent derives from the well-known practice of peer review. Experiments with online collaboration over the last few years have shown a willingness on the part of the public to offer the wealth of their personal knowledge to a cause they believe in.35 Patent examination is well-suited to pre-grant community participation because it depends on scientific expertise to make the correct determination. Peer-to-Patent aims to address the information deficit directly by bridging the gap between the community and the examiner.
The Peer-to-Patent pilot opens the patent examination process to public participation. Scheduled to run for one year (beginning June 15, 2007), it offers the first opportunity in the history of the USPTO to contribute prior art, commentary and suggested avenues for research directly to the Office and to get feedback from the patent examiner about the relevance of those submissions to the patent examination process. Although the patent examiner still controls the ultimate decision of patentability, this is the first time the general public has an opportunity to share information that, if relevant, can be used to narrow or even defeat the claims of a patent application. In essence, Peer-to-Patent allows the examiner to search the "human database" of people in the community who know something about that area of innovation.
The pilot is limited in scope to patent applications that are classified in USPTO Technology Center 2100 (TC2100) which covers computer architecture, software, and information security patents. While TC2100 is only 1 of 8 technology centers housed at the USPTO, it presents a perfect sample of applications to test the ability of peer review to deliver useful information. The innovation in the computer software industry is cutting-edge and much of the know-how is not contained in easy-to-find academic journals, making it the area of patentability that suffers most from the information deficit.
The sluggish pace of reform in this particular field can partially be attributed to the fact that the area of software patents is also among the more contentious. Many groups support eliminating software patents altogether36] but there is no indication that the USPTO intends to discontinue granting them.37 And, while the EPO is far more skeptical than its American counterpart when it comes to recognizing computer software as patentable subject matter, software patents from the EPO are not altogether unheard of.38 Thus it becomes increasingly important that those patents that are granted are a reflection of actual innovation occurring in the industry.
The rise of open source39 has made software prior art more widely available to the public but it is difficult for the USPTO to keep up. A significant amount of open source prior art remains unavailable via the normal searching methods employed by patent examiners. Furthermore, the speed at which the software industry moves, coupled with the average time it takes to get a patent, has made procuring a patent moot in many cases, which has no doubt left many inventions undocumented and therefore unavailable within the USPTO search systems. Without access to the relevant pool of knowledge, and with disclosure by patent applicants unreliable, patent examiners cannot correctly determine whether or not they should grant a patent. The examiner is unlikely to know about the open or closed source code, products or processes, websites or prior publications that ordinary people in the community know about from their personal experience.
Peer-to-Patent also provides an opportunity to demonstrate the importance of public accountability. If public review is institutionalized, it can help to ensure that inventors stop filing poorly drafted and egregiously unmeritorious applications and examiners stop granting them patents.40 Public scrutiny at an early stage in the patent lifecycle forces inventors to fully consider the consequences of filing an unmeritorious patent application, such as the impact of their actions upon their reputations.
It can also open up a conversation between the examiners and the public about the validity of software patents. Debate as to whether software is patentable subject matter abounds but there is very little in the way of providing a venue for funneling those ideas towards the people that are in a position to make changes. One of the most important distinctions between Peer-to-Patent and other sites that allow users to comment on patents is that Peer-to-Patent is run in cooperation with the USPTO. Examiners will actually receive user-submitted prior art. Providing relevant prior art to the USPTO that invalidates pending software patent applications is perhaps the best argument that opponents to software patenting can make.41 Successfully invalidating pending software patent applications will go a long ways towards beginning a conversation that is not easily avoided.
Still, opening the process for public accountability will depend upon the community's willingness to participate now in the pilot and demonstrate, to both the Patent Office and Congress, that the community - not just lawyers or patent professionals but technologists, students, hobbyists, engineers and others - has knowledge to contribute that can improve the process.
It is important to remember that patent examiners are dedicated to the integrity of patents, not granting patents. The examiner is empowered under the law to reject claims that are not novel or are non-obvious in light of prior publications. The system is designed to enable the public to be highly specific about the relevance of the prior art to the claims of the application. If any member of the community is able to show the Patent Office why, by law, the claims of an application should not issue, the USPTO is almost certain to welcome and use that input to strike the claims
Since its launch on June 15, 2007 (and as of this writing in August 2007), the Peer-to-Patent community has grown to over 1,400 peer reviewers.42 The value of creating open dialogue in patent law can be seen in the diversity of the community. Upon signing up, reviewers are asked to select what they consider to be their professional role. Possibly the most interesting piece of data is that although the Peer-to-Patent pilot is limited to applications concerning computer technologies, less than half (39%) of the peer reviewers classified themselves as a “Computer Professional/Technologist”.43 This seems to support the idea that members of the community, regardless of their actual area of practice, are knowledgeable and, perhaps most importantly, are willing to contribute.
This conclusion is further corroborated by the diversity of educational degrees among reviewers. While the majority hold engineering or computer science degrees, reviewers report having degrees in everything from comparative literature to applied mathematics, from the bachelors level through doctorates.44 This is significant considering that patent examiners are not required to possess an advanced degree.45 Having reviewers who hold masters degrees and higher assist in the examination process will itself be a noteworthy improvement to the existing system.
To date, the Peer-to-Patent site has received over 128,000 page-views from over 21,000 unique visitors in 109 countries, highlighting the transnational benefit that can be derived from peer review. Given the proposed PRA amendments to remove any dichotomy between prior art found domestically and abroad, it is obvious that foreign knowledge and expertise will become more and more valuable to the USPTO in their decision-making. Conversely, in an increasingly global society, people in any country can be harmed by patents granted as a result of an information deficit. During the time a non-meritorious patent is issued and is either invalidated by the courts or expires, the public may be forced to design around the patent claims, pay royalties, or otherwise comply with a threat that could have been avoided.
Many other countries are also experimenting with solutions. For example, in 2001, Australia introduced the “innovation patent”, which replaced the former “petty patent”. The objective of the innovation patent is to address the gap in patent protection for minor and incremental innovations. They are intended to be less expensive and quicker to receive than the standard patent.46 The period of protection for an innovation patent is 8 years, as opposed to the 20-year protection period granted under the standard patent.
Japan, recognizing the international importance of patents, has advanced an ambitious project called the Patent Prosecution Highway (PPH). The idea behind the PPH is to streamline the process for patent applicants to acquire foreign patents. Thus far, the JPO has launched PPH pilots with the USPTO, the UK IP Office, and the Korean IP Office. The program purports to reduce the workload of examiners by encouraging participating patent offices to utilize each other’s prior art searches and examination results.47]
There is merit to these programs but the problem is that none of these solutions address how to avoid granting low quality patents in the first place. The Australian innovation patent sheds 12 years of protection but the tradeoff appears to be a lower standard of patentability.48 In order to obtain a standard patent in Australia, an invention must be shown to take an “inventive step”, as opposed to the less stringent “innovative step” necessary for an innovation patent.49 Lowering the threshold for patentability is not likely to be the solution for granting higher quality patents, regardless of the term of protection. Indeed this merely expands the same problems associated with an information deficit to a new set of circumstances. The only difference is that instead of protecting a non-meritorious patent for 20 years, it would only be protected for 8. Moreover, the number of innovation patent applications filed is nearly the same as those filed for standard patents. Because examiners compare both types of patents against the same pool of prior art, deficiencies in information could potentially lead to a doubling of non-meritorious patents.
Japan’s PPH program promotes shared resources to expedite the examination process, something it has in common with Peer-to-Patent. The difference lies in timing. An application is not eligible for the PPH until it has already been deemed patentable by the office of first filing.50 This in itself presents difficulties when comparing the differences in patent ideologies between countries (for example the differences between the USPTO and the EPO with regard to patenting software as discussed above). Beyond that, if the consideration of patentability is made without using the best information, then there is a serious possibility that all that has been done is make a low quality patent easier to obtain on an international level. Furthermore, while the PPH does combine the resources of multiple patent offices51 to some improvement, the benefit is limited by the fact that all of the research systems are, to a degree, closed systems, whereas Peer-to-Patent actively seeks information that may not reside within the patent office systems.
Nothing in this article is meant to diminish the necessity of judicial and legislative responsibility in this field. In an article reviewing Peer-to-Patent in IP Law360, Mark L. Hogge rightly asserts that the majority of the patent lifecycle occurs after examination and that “[p]erhaps more focus by inventors and courts on the presentation and treatment of inventions would be helpful in preventing invalid patents.”52] But he appears to underestimate the reliance placed upon the patent examiner’s findings by inventors and courts alike and the consequences thereof.53 If that deference to examiners’ findings is to have any value, examiners must be provided with an encyclopedic pool of knowledge.54
Hogge further contends that there are better ways of stopping invalid patents, citing the creation of the Court of Appeals for the Federal Circuit as one.55 But Hogge fails to recognize that changing the venue for review does little, if anything to eliminate “invalid” patents. If anything this argument seems to cut against Hogge’s contention. Passing jurisdiction from a circuit court that traditionally invalidates patents to a court that is more likely to uphold patents makes it easier for a non-meritorious patent to live out its entire 20-year lifecycle despite a lack of legal merit.
Whatever the duty of the inventor or the courts to “do their jobs”,56 the importance of addressing the information deficit is highlighted by events like the KSR decision and the PRA. This is why Peer-to-Patent is so crucial. It is the first chance to get relevant information to the USPTO before the patent issues and not after it has become the subject of costly and protracted litigation. It is critical that these issues of patentability are resolved from the start rather than waiting for a judge who knows little about patent law and even less about science to make that determination.