WORLD TRADE CENTER SUITE 1800 401 EAST PRATT STREET BALTIMORE, MARYLAND 21202 410/659-9550 FAX: 410/659-9549 larry@GuffeyLaw.com
"Maximizing the Returns on Your Intellectual Endeavors"

PATENT PROTECTION FOR COMPUTER SOFTWARE,  May 1998 MD Bar Journal, pp. 54-57.

            The founders of the United States provided in the Constitution that "Congress shall have the power...to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries," U.S. Const. Art. 1, § 8, cl. 8.  A patent is the grant from the United States government to the invention’s owner of the right to exclude all others from making, using or selling the invention for a limited time.  In exchange for these patent rights, the inventor is required to file and have approved a patent application which discloses the invention in enough detail to allow other people to use the invention after the monopoly expires and to learn from the invention and build on its ideas.

            There are different types of patents (i.e., utility, design and plant). A utility patent protects the functional aspects of an invention and is the type of patent that is applicable to the protection of computer software.  A utility patent protects the invention for a period of twenty years from the date of filing of the required patent application, subject of the payment of specified maintenance fees.

Complimentary Nature of Patent Protection

            Patent protection has recently become a more realistic, valuable, additional means, complimentary to traditional copyright and trade secret means, for protecting computer software.  The truth of this statement  can be seen by considering the following infringement hypotheticals and assessing which means of protection are most applicable and useful.

            Assume that an inventor: (a) conceives of and reduces to practice a new, nonobvious and useful piece of software, (b) hires an independent contractor, computer programmer to write the code expressing the invention, (c) contracts with a software publisher to produce the resulting software product, (d) contracts with a distributor to distribute copies of the software, and an end-user purchases a copy of the software.

            Noting that in this situation a patent, copyright and trade secret on the software will respectively protect its functional aspects, its specific manner of expression and those not-generally-known details of the software that are not revealed in the patent, what probable recourse is available to the software inventor with these types of protection in the following situations?

            (1)  An independent party, who has not been involved with the other parties and has never seen the software product, just happens to create, manufacture and sell a product that is nearly identical:

Patent -                       Damages available, even independent creators are infringers.

Copyright -                  Nothing, no access to the product, can’t prove copying.

Trade Secret -             Nothing, no relationship with others, a requirement for protection.

            (2) The software programmer goes out and resells the source code to another software publisher, and the inventor only had a handshake deal with the programmer to write the code.

Patent -                       Damages available, selling of the invention is an infringing act.

Copyright -                 Nothing.  The programmer actually owns the copyright rights since the inventor didn’t make the necessary arrangements to have the programmer assign these rights to the inventor.

Trade Secret -            Damages available, have necessary relationship.

            (3)  An end-user makes copies and passes them to friends.

Patent -                       Damages available, but it’s not economically reasonable to go after this infringer.

Copyright -                 Assuming assignment of the programmer’s rights, damages available. Lower cost method may exist for enforcing these rights; the Software Publishers Association may assist with enforcement of rights.

Trade Secret-             Nothing, no relationship with others, a requirement for protection.

            (4) The software publisher creates its own version by hiring a team of programmers to employ a standard, copyright-infringement-avoidance “clean room technique” (i.e., group of “dirty” programmers study the product and express how it works in terms of functional diagrams that are given to a group of “clean programmers”  who, never having had access to the product, proceed to write the necessary code for the new product) to make a competitive product that is then offered for sale.

Patent -                       Damages available, selling functional equivalent of the invention.

Copyright -                  Nothing, no access to the product, probably can’t prove copying.

Trade Secret -             Damages may be available if not-generally-known details of the software that are not revealed in the patent are used; have the necessary relationship.

            Patent protection is always seen to provide the inventor with a remedy.  However, the use of patent protection has its potential drawbacks: (a) the acquisition of one’s patents rights in the invention is often difficult, time-consuming (i.e., 2 - 3 years) and relatively expensive, and (b) the cost of enforcing one’s patent rights is expensive (i.e., $50,000 typical litigation retainer). 

            Because of its wide range of applicability, patent protection compliments the traditional copyright and trade secret means of protection for computer software.  Over-lapping these means of protection usually provides the inventor with the most effective protection.

 Recent Increase In The Importance Of Patent Protection

            Recent events have combined to increase the importance of patent protection: (1) copyright protection has been narrowed, (2) patent protection for software and related processor-based systems has been expanded, and (3) new guidelines for the examination of computer-related inventions, which are expected to yield fewer patentability rejections on the grounds that the claimed invention is non-statutory subject matter, were issued on March 29, 1996 by the U.S. Patent and Trademark Office (USPTO).

            In Computer Associates v. Altai, 982 F.2d 693 (2d Cir. 1992), the court held that other than the copying of the literal elements of a computer program (i.e., the object codes and the source codes), copyright offered relatively narrow protection for the other elements of software (i.e., the structure, sequence and organization of the software).  In 1996, the Supreme Court affirmed the holding in Lotus v. Borland, 49 F.3d 807 (1st Cir. 1995), that the command menu hierarchy of the Lotus 1-2-3 spreadsheet program is not copyrightable because it is a method of operation.  In both of these cases, the opinions suggest that what the plaintiffs were seeking was protection that falls within the arena of patents -- not copyrights.

            Meanwhile, in 1994 the Court of Appeals for the Federal Circuit (CAFC) decided five cases involving the question of whether software inventions constituted patentable, statutory subject matter under 35 U.S.C. § 101.  These cases generally tend to broaden the scope of pantentability for software inventions, and indicate heightened interest in this question (i.e., In the twelve prior years since the CAFC’s creation in 1982, the CAFC has only heard three similar cases).

            In In re Alappat, 33 F.3d 1526 (en banc), the CAFC held that a means for creating a smooth waveform display in a digital oscilloscope was patentable subject matter.  Thus, a “general purpose” computer operating pursuant to a computer program to perform specific functions was held to be a discrete, patentable machine.

            In re Lowry, 32 F.3d 1579, did not directly present a statutory subject matter rejection issue to the CAFC.  This issue was resolved earlier by the USPTO Board of Patent Appeals that reversed the patent examiner’s rejection of a claimed means for  imposing a physical organization on data that provided computational advantages for the storage, use and management of information residing in a computer’s memory.  The CAFC went on to hold that the USPTO must consider all the claim limitations (of this invention) when assessing the invention with respect to the novelty requirement (35 U.S.C. § 102) for patentability.  

            In In re Schrader, 22 F.3d 290, the CAFC held a method of bidding at an auction to be unpatentable, noting that “there is nothing physical (i.e., connected with the physical environment - a transformation or reduction of matter) about the bids per se.”  However, a dissent argued for patentability on the basis that the method “achieves a technologically useful result.”

            In In re Warmerdam, 33 F.3d 1354, the CAFC held a method for controlling the motion of objects such as robots to avoid collisions with other objects to be unpatentable because it describes “nothing more than the manipulation of basic mathematical constructs, the paradigmatic abstract idea.”  However, it did hold a claim in this same patent for a computer storing the computational results from the application of the method to be patentable.

            In In re Trovato, 60 F.3d 805 (1995, en banc), the CAFC vacated the 1994 judgment of a three-judge panel that a linear programming technique to find the optimal path between two locations was non-statutory subject matter and remanded the issue to the USPTO Board of Patent Appeals.  The CAFC noted that In re Alappat was decided subsequent to the three-judge panel’s decision and provided further guidance on the 35 U.S.C. § 101 statutory, subject-matter issue.  Furthermore, the CAFC noted that the USPTO had recently issued “Proposed Examination Guidelines for Computer-Implemented Inventions” and ordered that the case be reconsidered in light of any new guidelines to be adopted by the USPTO.  

Patentability of Computer Software

            All inventions must satisfy the following patentability requirements: (1) useful, patentable subject matter, 35 U.S.C. § 101; (2) novelty, 35 U.S.C. § 102; (3) nonobviousness, 35 U.S.C. § 103.

            35 U.S.C. § 101 is broad and general in its language describing patentable subject matter:  “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.”  Patentable, statutory  subject matter is seen to be either a useful “process,” “machine,” “article of manufacture” or “composition of matter.”  Thus, this statute limits patent protection to inventions that possess a certain level of real world value as opposed to those that just represent nothing more than an idea or concept.

            The new USPTO guidelines for the examination of inventions related to computer software (See Manual of Patent Examining Procedure §2106 - Patentable Subject Matter, Computer-Related Inventions) indicate that there are three general ways whereby software can be used in an invention to satisfy the statutory subject matter requirement.  First, software can be used to create a “machine” when a computer is included in the claims as having its operation controlled by the software, as illustrated in In re Alappat.        Second, software can be claimed as a “process” when it is described as a series of operational steps to be performed on or with the aid of a computer.  And finally, a computer-readable memory (e.g., floppy disks, CD-ROM, system memory) containing the software that can be used to direct a computer to function in a particular manner can be considered as an “article of manufacture.”

            The novelty requirement for an invention will be satisfied unless, during the patentability examination of the inventor’s application seeking the patent rights for the invention, a USPTO patent examiner can cite a single prior art reference from anywhere in the world that contains all of the features claimed in the invention disclosed in the inventor’s patent application.  Attempts are made to avoid such novelty rejections by carefully drafting the claims within the patent application so that they contain new features or elements not revealed in the prior art.

            The probably more demanding, nonobviousness requirement is harder to objectively define.  The typical way that an examiner shows obviousness is to cite a number of prior art references that, when combined as suggested by possibly another prior art reference, contain all of the elements of the applicant’s invention.  The applicant generally challenges the logic of the combination and identifies certain secondary considerations that suggest that the invention was not obvious (e.g., long duration of well-recognized need for the invention, prior unsuccessful efforts of others to develop such an invention, degree of commercial success of the invention).

            Finally, the Supreme Court has long held that certain types of invention are unpatentable: (a) laws of nature or science, (b) naturally occurring articles or natural phenomena, and (c) abstract ideas.  In Gottschalk v. Benson, 409 U.S. 63 (1972), the Court extended the definition of law of nature to include “algorithm, or mathematical formula.”

            Since computer software often uses algorithms and mathematical formulas, this holding raised many doubts as to the patentability of computer software.  However, in Diamond v. Diehr, 450 U.S. 175, 186 (1981) the Supreme Court clarified that an unpatentable algorithm means “a procedure for solving a given type of mathematical problem.”  Thus, computer software containing a mathematical algorithm and claiming something other than just the algorithm itself is now recognized as being patentable if it can meet the other statutory requirements for patentability. 

 Process of Obtaining a Patent

             Because of the often complex and specialized nature of the patent application process, most inventors chose to retain the services of a patent attorney (i.e., an attorney admitted to file and prosecute patent applications before the USPTO) to represent them in this process.  Since the legal expenses associated with applying for a patent can be relatively large (i.e., $3,000 to greater than $10,000), it is often advisable to first make a relatively low-cost assessment (i.e., a novelty search at a cost of $500 to $2,000) as to whether it is likely that a patent will issue on the proposed invention.  This typically involves investigating USPTO files for technology similar to the proposed invention to determine the state of the art in the area of the invention and what, if any, scope of patent protection may be available for the invention.  If the search reveals no prior patents that substantially disclose the proposed invention, this knowledge serves to justify incurring the patent application preparation expenses.

            The patent application has two primary sections: the specification and the claims. The specification is essentially a detailed description of the invention, including drawings if necessary, showing what the invention does, how it works, and disclosing its advantages over prior art. The description of the invention must meet certain legal requirements, including that it: (a) be thorough enough to allow a person skilled in the invention’s field of technology to make and use the invention, (b) identify the invention’s best mode of operation, and (c) provide a basis for and explanation of the terminology that is used in the following claims.

            The claims of the patent application are carefully worded, legally structured sentences that vary in definitiveness, with each of these sentences attempting to distinctly define more closely than the others the patentable novelty of the invention.  Thus, while one ideal claim might suffice, the patent application almost always contains claims that range from broad ones (which define the invention with the fewest details possible) to narrow ones (which set forth in more detail the specific elements of the invention).

            For the USPTO to examine a patent application, a required filing fee of  $790 (half of this amount for independent inventors and companies with less than 500 employees) must be paid at the time the application is submitted.  After the patent application has been filed with the USPTO, a patent examiner carefully reviews the application in order to determine the invention’s patentability. The examination workload and staffing of the USPTO are such that this examination process usually extends over a period of 18 to 30 months.

            When the examiner has made an initial patentability determination, the USPTO sends the applicant an “Office Action” containing its determination and the reasons for it.  If the Office Action contains a rejection of the claims (which occurs on the first Office Action about 80% of the time) and there exist arguable grounds for contesting the examiner’s determination, one typically files a “Response,” usually in the form of an Amendment, to overcome the rejection. There are usually only two Office Actions and Responses before a final determination is made by the examiner as to the invention’s patentability.  If this determination is detrimental to the applicant’s interests, it can be appealed in the federal court system.  The courts ultimately decide the validity and enforceability of a patent.

Impact of the New Guidelines for Examining Computer-Related Inventions

            In the thirty years prior to the new guidelines, the courts often used various formalistic grounds for rejecting computer software claims based on lack of 35 U.S.C. § 101 subject matter (e.g., the claims entailed merely mathematical algorithms, printed matter, methods of doing business, or mental steps).  The new guidelines have abandoned such grounds.  In their place, the USPTO has reinstated the three basic classes of non-statutory subject matter recognized by the Supreme Court prior to Gottschald v. Benson: (a) laws of nature or science, (b) naturally occurring articles or natural phenomena, and (c) abstract ideas.

            In addition, the new guidelines focus on the claimed technological utility of the invention.  This technological utility requirement is in addition to the more generalized 35 U.S.C. § 101 requirement that an invention be “useful,” which has usually been interpreted as requiring only that an invention not be: (a) wholly inoperative (i.e., inventions that will not operate as described), (b) violative of basic scientific principles (e.g., perpetual motion machines), and (c) lacking in socially beneficial results (i.e., inventions that are illegal, immoral or otherwise contrary to public policy).

            The new guidelines also emphasize the classification of each claim under examination as to its 35 U.S.C. § 101 subject-matter category (i.e., article of manufacture, machine or process).  If the claims presented do not fit neatly within a  category, they will be rejected, even if a statutory invention is disclosed in the specification.   This situation emphasizes the importance of drafting computer software claims so that the claimed invention clearly falls within one of the acceptable subject- matter categories.

            However, the new guidelines also note that a claim which appears to be within a statutory subject-matter category based on its form, may actually define non-statutory subject matter (i.e., abstract ideas, laws of nature and naturally occurring articles).  An example given in the guidelines that represents this situation is: “a method consisting solely of the steps necessary to converting one set of numbers to another without reciting any computer implemented steps.”  Thus, to avoid a non-statutory, subject-matter rejection when claiming software for a purely computational algorithm, it would appear that the claim should be drafted in terms of the internal operations of the computer when it is executing the software.  In summary, the new guidelines' main substantive contribution is to shift the focus of the subject-matter inquiry from content-specific, mechanical rules, to subject matter classification in terms of the 35 U.S.C. § 101 statutory categories and technological utility.  The fact that a claim is cast in terms of a statutory category is not sufficient; the specific utility of the invention must also be apparent from the claim.  On the other hand, under the new guidelines it is not necessary for the subject matter to interact with the physical environment external to the computer (as was required in In re Schrader); although, if it does, this probably increases the likelihood that the subject matter will be found to be statutory.  Rather, the issue is whether the subject matter produces a sufficient "tangible benefit" so that it cannot be viewed as merely an abstract idea (a position closer to the dissent in In re Schrader). The net result of the new guidelines should be that patents directed to computer software inventions will more readily issue, and those that do issue will have an increased likelihood of withstanding statutory subject matter challenges in litigation.