First Panel: The Great Debates: Software Patents and Patent Reform
Brian Kahin, Senior Fellow at the Computer & Communications Industry Association and Adjunct Professor at the University of Michigan School of Information, opened the day with a review of the problems with software patents. He talked about companies that acquire a portfolios of software patents, as when a firm either applies for or purchases a multitude of patents that do not necessarily cover the technology that they are developing. Most software firms do this to defend themselves against litigation from companies that are looking to build a profit off of their own patent portfolios.
James Heald, a member of the Foundation for a Free Information Infrastructure (FFII) and their United Kingdoms coordinator from 2003 to 2005, presented the current situation of software patents in the European Union. He reported that to date there has been no patent litigation and that the judicial system is fractured on software patents. The European Patent Office's definition of patentable software is vague, requiring what Heald described as "technical solutions to a technical problem," to chuckles from the crowd. With poor access to the courts and no mention of patents in the EU's constitution, Heald is worried about the unclear future of European software patents. Most software firms, especially smaller ones, would rather see the industry run by copyrights.
The first panel concluded with Robert Barr, executive director of the Berkeley Center for Law & Technology. Most law groups software and hardware together, using legislation for the semiconductor industry to regulate the software industry, Barr said. This use fails when many hardware manufacturers also develop software, such as Cisco Systems implementing its own operating systems on its network equipment (Barr was previously the vice president for intellectual property and worldwide patent counsel to Cisco Systems). Barr then cited hopeful patent reform that is in discussion, which uses damage apportionment and willful infringement. However, he is skeptical as to whether the agenda will work.
Second Panel: Empirical Evidence on Software Patents
The second panel began with Bronwyn Hall, an economist and professor at both the University of California Berkeley Graduate School and Economics of Technology and Innovation at the University of Maastricht, who presented an in-depth analysis using economic mathematics. She presented an event study that looked at how 1994 and 1995 patent legislation affected the software industry and whether software patents are valuable to firms. Her data was taken from publicly traded firms from several non-software industries, allowing for a comparative analysis. Hall then looked at the five-day market effect on hardware and software firms from the In re Alappat 33 F.3d 1526 (Fed. Cir. 1994), In re Warmerdam 31 USPQ2d 1754 (Fed. Cir. 1994), In Re Lowry 32 F.3d 1579 (Fed. Cir. 1994), and In re Beauregard 53 F.3d 1583 (Fed. Cir. 1995) and United States Patent and Trademark Office (USPTO) guidelines. She concluded that the majority of software patents are not valuable, and yet there has been an increase in their numbers.
Jim Bessen, a lecturer at Boston University School of Law, brought a software developer's perspective. Bessen showed statistics on the likelihood a firm would be involved in a patent suit by industry. He asserted that the cost of patent litigation is greater than the profit from it.
Third Panel: Effects of Software Patents on the FOSS community, entrepreneurs, standards, information and communications, ecommerce, and financial services
Representing the FOSS community, Eric von Hippel, professor and head of the Technological Innovation and Entrepreneurship Group at the MIT Sloan School of Management, reviewed how a distributed development network generates technology, and how people assign pay or credit for inventions. Using PostgreSQL v7.4's intelligent data vacuuming as an example, von Hippel calculated how many individuals were involved in the feature's design and development, and then pointed out how it is impossible to assign a patent for the implementation. "The innovators do not file patents for what they developed, but the firms that they worked for do."
Pamela Samuelson, director of the Berkeley Center for Law and Technology and an Electronic Frontier Foundation (EFF) board member, was representing entrepreneurs. Focusing on the startup stage, Samuelson reported that most entrepreneurs do not consider patents or how they could be affected by them. Startup firms do not consider patents because they do not have the capital for patent clearance with the USPTO and they are still small enough that they do not feel threatened by patent trolls. However, this means that when they have an initial public offering of stock, they open themselves to increased troll attacks. Like many presenters, Samuelson calls for a better patent application review process.
Next was Daniel Weitzner, director of W3C's Technology and Society Domain, who spoke about standards. Weitzner pointed out that at W3C's beginning they did not consider patents, as they were more concerned with the technology. W3C did not consider standards for licensing technology; it did not occur to them to do so, as that was not how developers had acted previously. The Internet had always been controlled by commonly accepted standards. The problem with getting all developers to agree on one standard of licensing is the number and dispersal of those developers.
Covering information and communications, Jason Shultz, a staff attorney leading EFF's Patent Busting Project, focused on software patents that prohibit freedom of speech. Patent portfolios create a monopoly and inhibit innovation, he said, citing the 200 applications for patents on blogs and innumerable existing patents on methods of accessing data and social interaction. Shultz cited the difference between software and non-software patents, such as the telephone: with software, "the role of the people was novel, not the technology." Shultz continued by claiming that software patent holders are no longer going after other firms, but against end users. For example, if some of those blog patents are granted, then Google, which purchased Blogger, could be open to patent infringement litigation.
Representing ecommerce, Emily Ward, associate general counsel on patents to eBay, provided her firm's experiences with software patents. "There is no reason to treat software differently," she asserted, but the software industry makes firms large targets for infringement litigation because of the ease of acquiring a software patent. Most large firms do not try to enforce their patents, because it would not be profitable. If they are facing another large firm, which likely also has a large patent portfolio, then there may be no winner, just as there is theoretically no winner in a nuclear war. However, smaller firms do not have these same concerns. Often, when a firm is going into the red, it will try to use its patents as assets to raise capital through lawsuits. Ward emphasized this by stating that eBay gets approximately 100 filings against the company per year for patents. Those patent trolls will often file patent litigation only in certain districts, knowing which ones are more likely to invalidate a patent.
Rounding out the panel was Bob Hunt, senior economist from the research department of the Federal Reserve Bank of Philadelphia. Hunt began by showing how financial services are connected with software, in the implementation of statistical models, connecting data (such as credit information), and hedging technology. Looking more generally at software patents, he said a "patents arms race" only generates more patents without increasing research and development in firms; instead, Hunt would like to see patents be less broad and more closely related to the technology. he noted that further damage can be caused when a non-expert in the industry files a patent, which may force the market in a certain direction by closing possible routes of innovation.
Fourth Panel: Legal Perspectives on Software Patents
John Duffy, a professor at George Washington University Law School, started the fourth discussion by claiming that the large differences between the software industry's and alternative industries' patents are not as large as many had suggested that day. All industries have fuzzy boundaries, multiple innovation products, small firms, and distributed innovation. Furthermore, the passing of broad patents by the USPTO is not purely a software industry problem. Duffy would like to see tougher obviousness standards across the entire patent system.
Peter Menell, professor at University California Berkeley School of Law, covered US patent legislation. For any software or business patent to pass, it must be a "process" under 35 U.S.C. 101 and "anything under the sun made by man" under Senate Report No. 1979, 82d Cong., 2d Sess. as instituted in 1952. Under the legislation, "process" means that the product must produce a physical effect, which does not cover software. The Title 101 wording was put forth in 1952, before the software industry came into being, and its authors would not have considered software to be covered by its wording. Secondly, using "anything under the sun" to pass broad patent applications is exploitive, as the wording outside of the ellipses requires adherence to the other titles as well as 101, "anything under the sun that is made by man, but it is not necessarily patentable under section 101 unless the conditions of the title are fulfilled." Menell wants the courts and patent system to look outside of the ellipses and consider the historical flaws.
Michael Meurer, professor at Boston University Law School, opened his discussion with the taxonomy of patents, which turned into a list of problems: fuzzy boundaries, public access to patent information, property rights not being tethered to possession, and the cost of patent clearance. He related the property rights and possession issues with an anecdote from when his father would go duck hunting -- "There was a common understanding that when he shot and killed a duck that he did not claim ownership of the whole flock." Disagreeing with Duffy, Meurer insisted that patent problems are more prominent in the software industry. Currently the USPTO is discussing reforms that should fix some of these problems, but there is not enough lobbying for changes. IBM has suggested more transparency in patents, allowing the inventor to be on file instead of only the firm that filed the patent.
Jay Draftler, professor at the University of Akron School of Law, began his talk by comparing how chemical and software patents are treated. When a chemical equation is patented, only the catalyst is patentable, not all of the variations of chemicals in the equation. However, in software, the entire flowchart is patentable, which is broader in scope than in any other industry. Focusing on patent applications, Draftler looked at the language of "whoever claims" in Title 101 and whether it applies to software. While asserting that it does not, he would still like more focus to be kept on this language.
Draftler suggested that software development does not constitute invention. In order for something to be an invention, it requires technical risk and therefore the risk of failure. Software does not have such a risk. Therefore, the copyright system is more applicable to software than patents. Patents' failures, Draftler said, are demonstrated by IBM, the number one patentee for the last 30 years, dumping large sections of their patent portfolio into the public domain.
Fifth Panel: Roundtable: Options for action/reform
Dan Bricklin, president of Software Garden, began the fifth panel by discussing the problems caused by uninformed juries during patent appeal hearings. It is conceivable that none of the jurors have used a defendant's technology or understand how it relates to the plaintiff's patent. Bricklin would like to see at least one person who uses said technology on every jury, and for the jury's pay to be increased to that of the lawyers to increase their investment in the case. Patent appeals are so important because they can affect the entire industry the patent is in, acting as a "super precedent."
Solveig Singleton, Senior Adjunct Fellow at the Progress and Freedom Foundation, suggested that there are too many possible solutions to the heaps of problems posed by software patents, ranging from removing the patent system altogether to improving the patent reviewer institutions. Whatever the solution, she said, it needs to focus on business incentives and reforming the patent application review process. The reason why the USPTO is not so concerned with keeping itself in check is it does not have a large incentive to do so, compared to the military, for instance, where people trying to kill soldiers acts as an incentive for those involved to do their best.
Ben Klemens, guest scholar at the Brookings Institution, focused on the broadness of software patents. His primary example was that Barnes and Nobles is open to litigation from Amazon due to their online presence. The broadness of so many software patents opens any non-software firm that has an IT department to patent infringement. From this situation, Klemens pulls his own definition of a troll as any informed group that claims patent infringement on a misinformed group. Klemens puts his faith in the judicial system rather than in Congress for a solution, "Now that we have lawyers pissed off about patents, we'll see something happen."
Ed Black, president and CEO of the Computer & Communications Industry Association, spoke about the problems anyone might encounter when attempting to reform the USPTO. One of the largest problems he cited was that "only about 12 congressmen actually understand this technology." Black would like to see someone publish a series of one-page papers that someone with no previous knowledge could read in the elevator and come out understanding the problem and what needs to be done. A hurdle to this end is that patent lawyers, who Black feels are more concerened with profits than anything else, have previously been the more eloquent camp.
Finally, Richard Fontana, counsel to the Software Freedom Law Center, talked to the relationships between open source and the USPTO. Wanting to see more disclosure in patent applications, Fontana suggested required deposits of source code similar to biological deposits required in other industries. Fontanta concluded that open source does not have much to worry about from software patents, saying that "looking at open source for infringement is not realistic." Nor would open source take any larger role in software patents until trolls and broad patents are abolished. However, while the likelihood of a suit is low, open source generally protects itself by adding anti-patent clauses to its copyright licenses, such as in the GPL v3 (7/27/2006 draft) and v2 (June, 1991).
While certain parts of the conference felt like review, it was pleasing to see so many academics from such a broad range of backgrounds engaging each other and playing devil's advocate, though I wish the sessions had had more lively Q&A sessions. Nevertheless, it is clear from the conference that some form of change must be made, whether from the judicial or legislative branch, USPTO, or a grass roots movement. Most likely it will take a grass roots movement of patent lawyers in tandem with engineers and developers to see any changes in the near future.