Strategies for Litigating Computer Software Copyright Claims

Article originally appeared in Litigation Magazine on 10/01/07

If a client came to you and told you that his novel had been stolen and published under the name of another person, the path to litigating your case would be fairly evident. You would take your client’s book, along with any drafts, notes, or other evidence of when she wrote it and compare it to the book published by the alleged thief. You would look for line-by-line copying and for copying of the narrative themes and characters. If the book was published in another language, you would hire a translator to compare your client’s version to the published work to find any similarities indicating copying. Your lifelong familiarity with printed fiction as well as nonfiction would make litigating the case straightforward for you and for the jury.

Computer software is a different matter altogether. Although governed by the same provisions in the Copyright Act as other literary works, copyright cases involving computer software present special challenges. Unlike copyright cases involving works we are familiar with—say a book, a song, or a painting, where we as litigators start with at least a superficial familiarity of the medium upon which we can begin our assessment of the case and preparation of our litigation strategy—software cases often present the challenge of how to litigate the unfamiliar.

An initial hurdle for the litigator facing a software copyright infringement case is to understand the precise rights at issue. Software cases often involve rights that fall within the realm of different laws, including patent law, copyright law, and trade secrets law, among others. Differentiating among these laws and their corresponding rights and protections is essential.

Generally speaking, patent law protects inventions, copyright law protects creative expression, and trade secrets law protects confidential business information. As applied to computer software, these lines are easily blurred. Under the patent laws, a computer program is patentable if it is novel and accomplishes a function that has some practical application. Patentability lies in the function accomplished by the software rather than the elegance of how the programmers expressed that functionality in the software. Copyright protection, in contrast, extends only to the programmers’ original, creative expression and not to the functionality imparted by the software. This distinction can be a fine line when dealing with computer software because software, unlike other creative works such as literature or music, typically serves a functional purpose. That functionality, however, is not protected by the Copyright Act. The processes and algorithms embodied in the software might be patentable, but they cannot be copyrighted. Copyright law protects only the expression used to accomplish the functionality, not the underlying functionality itself. E.g., Apple Computer, Inc. v.Microsoft Corp., 35 F.3d 1435 (9th Cir. 1994).

Trade secrets law, in turn, differs from both patent and copyright laws. Trade secrets protection turns not on the creativity or functionality of the information for which protection is sought but rather on whether the information is valuable to a company because it is not known to its competitors. Importantly, a computer program can be eligible for patent, copyright, and trade secrets protection, although the protection afforded by each of these different areas of law is unique and specific.

This article focuses on the rights afforded by the copyright laws. Under the Copyright Act, a “computer program” is defined as “a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result.” 17U.S.C. § 101. By its terms, the CopyrightAct extends to the literal set of instructions used to create the program. In addition, protection can extend to nonliteral components of the computer program such as its structure and

All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. organization. And protection can also extend to any audiovisual displays generated by the program, even if the underlying instructions used to create the displays are not themselves copyrightable.

One of the most fundamental tasks for the litigator is to identify the particular elements of expression embodied in, and generated by, the computer program for which copyright protection is sought, and then ascertain for each element whether it is copyrightable. Differentiating the protectable elements from the unprotectable is difficult and will often be one of the decisive points of contention in the lawsuit. One threshold approach is to assess whether the functionality performed by the software is capable of being expressed in multiple ways. The idea is that, if the source code used to perform a function can be written in multiple ways, copyright should protect the expression chosen by the programmer to accomplish the function. Conversely, if technological or other constraints limit the number of ways to express the functionality, or if the expression arises from the use of common ideas or their logical expressions, the instructions used by the programmer may not be subject to copyright protection. Computers are binary, that is, they can only “understand” 0s and 1s. People, on the other hand, cannot easily make sense of the 0s and 1s executed by a computer. This apparent disconnect was long ago resolved through the creation of programming languages. Today, most programmers write computer programs in what are known as high-level programming languages such as Java, C, Basic, Fortran, and Cobol. Programming languages allow programmers to create software using text-based instructions that can be easily understood by other programmers. Computer code written in a programming language is known as “source code.” Programs written in source code are then converted, typically through the process known as compiling, into the 0s and 1s that can be executed by a computer. The resulting machine executable code is known as “object code.” Source code and object code are viewed under the copyright laws as two representations of the same computer program. Copyright protection extends to the computer program and encompasses both the source code and the object code. If the software at issue is in fact copyrightable, the copyright holder must consider the appropriate measures to protect her creation. The exclusive rights of the author/programmer under the Copyright Act vest the moment the software is created, because it is the act of creating an original work that gives rise to copyright ownership. 17 U.S.C. § 201(a). Those exclusive rights include the right to copy what has been created as well as the right to create modified, amended, or translated versions of the original creation—what are called “derivative works.” 17 U.S.C. §§ 101, 106. The owner can also license the software to others and even transfer her rights (in whole or part) through a written instrument. Registration of a copyright is not necessary to give the author any of these rights. Registration of the copyright, however, is required for litigation and is a prerequisite for recovery of certain remedies. 17 U.S.C. §§ 411, 412. Registration of the computer source code also creates a presumption of validity of the copyright in the software. Nonetheless, copyright registration is something of a mixed bag for the software author. Software is unusual among the types of creative works that may be copyrighted in that it can be (and usually is) distributed for use by others without revealing the content of the underlying creative work. When you purchase software at your local office supply store, what you get is just a copy of the object code, not the underlying source code. The object code, unlike the source code, does not reveal the creative expression of the program’s author.Although object code can be “decompiled” from its 0s and 1s into a high-level programming language, the resulting code can be expected to differ quite dramatically from the source code written by the program’s author. Consequently, by distributing only the object code while retaining the source code as a trade secret, a software author can make and sell many copies of her protected work without ever revealing the contents of her creative expression to potential infringers. Because of the ability to distribute object code without revealing the program’s source code, whether to register a copyright in that software becomes a strategic choice. In order to register a copyright in any work of expression, the author must complete an application describing the work and the circumstances of its creation and must deposit copies of the work with the Library of Congress. 17 U.S.C. §§ 407- 409. Once the work is deposited with the library, it becomes a public record that can be accessed by anyone. The deposit rules provide special protections for computer code, including allowing deposit of excerpts of the source code in lieu of the entire program—generally the first and last 25 pages of the source code, or the entire source code if the program is 50 pages or fewer. 37 C.F.R. § 202.20(c)(vii). Nevertheless, many programmers fear even a limited disclosure of their source code so much that they resist registering copyrights in their source code until necessary for purposes of litigation. The Copyright Office also allows deposit of the object code in lieu of the source code, but this option comes with a considerable price: The registration is given no presumption of validity because the deposit of object code does not allow the Copyright Office to determine whether the registered work contains copyrightable material. See id.; see also Syntek Semiconductor Co., Ltd. v. Microchip Technology Inc., 307 F.3d 775, 779-80 (9th Cir. 2002). Of course, if the copyright author chooses to postpone registration of the copyright, she may still protect her source code as a trade secret and may take other steps to protect her interests, such as the use of nondisclosure agreements. Identifying the author of the copyright material may not be as simple as it seems. Computer software is often the product of multiple programmers.Maintaining ownership over the work therefore requires particular care.Without an agreement to the contrary, the CopyrightAct provides that software created by employees acting in the course and scope of their employment is owned by the employer, whereas the copyright in software created by independent contractors is owned by the independent contractors unless they have agreed in writing that the software is a “work

Written agreements are a prudent measure to avoid surprises about who owns what.

All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. for hire” owned by the party paying them to write it. 17 U.S.C. § 201(b); Community for Creative Non-Violence v. Reid, 490 U.S. 730, 742-43 (1989). Ownership is often at issue in litigation. Written agreements are a prudent measure to avoid surprises about who owns what and to pin down the employment status of contributing authors. Written agreements are especially crucial if contract programmers are involved in the creation of the software, in which case agreements should be in place that specify what ownership, if any, the programmers will obtain in the software they create. Written agreements are also important when hiring another company to write custom software. The agreements should make clear whether your client will acquire a license, an exclusive license, or ownership of the custom code being created for it. Of course, the hiring company should ensure that the company writing the software has agreements in place with its own programmers to ensure that the company, rather than the programmers, owns and can assign the rights in the software. The lawyer faced with litigating a software copyright case should be conscious of software-related issues from the outset. Before any confidential source code is produced, a confidentiality order should be in place to protect the client’s valuable trade secrets. Often, such confidentiality orders restrict review of the produced source code to only trial counsel and independent experts. E.g., Brown Bag Software v. Symantec Corp., 960 F.2d 1465 (9th Cir. 1992) (restricting in-house counsel from reviewing trade secret information produced in discovery). Before jumping into a restrictive confidentiality order, however, consider the ramifications. Often you can obtain tremendous benefit by having someone who is familiar with your client’s software, such as your client’s chief programmer, analyze the opposing party’s software for similarities. The tradeoff, of course, is that for this to happen, you will have to agree to reciprocal analysis of your client’s software by her adversary— something to which your client will likely be loathe to agree. And disclosure of your client’s source code to her adversary could destroy trade secrets protection over her software. Our suggestion is that you consider this decision carefully. The quick decision is typically to prevent all party representatives from examining the other side’s software, but our experience is that at some point in the case you will really want your client’s assistance to analyze the key copyright issues of substantial similarity and protectable expression. You should balance that possibility against the need to protect your client’s own secrets. The new federal electronic discovery rules have raised the prominence of e-discovery in almost every case. But electronic discovery takes on additional importance in the software copyright case. In other types of civil litigation, electronic evidence is one of many types of evidence that might reveal relevant information about the parties’ claims and defenses, even though the claims themselves do not require that evidence be in an electronic form. In the software copyright case, the claims are usually focused on creation and ownership of electronic information, and the key evidence in the case is almost certain to be in electronic form. Perhaps the most basic evidence in a software case will be the parties’ respective source codes. You will need to obtain the source code to the software at issue in electronic form because that is really the only form useful for thorough expert analysis. Additionally, the computers on which the software was written, the computers on which it has been stored, and electronic copies of the software that can be found on any media (e.g., CDs, floppy disks, flash drives) should also be forensically inspected by experts for information about the circumstances in which the software was written, revised, or copied. Metadata is especially critical evidence in the software copyright case. Examination of the computers used to create the software at issue, and their storage media, can reveal creation dates, modification dates, and the contents of different iterations of the software.

Another source of useful information is provided by “version control systems” that are commonly used by programmers to track changes made to software both during its initial development phase and later during revisions and updates. These systems, which are akin to the document management programs that many lawyers use to store and share documents within their firms, provide valuable information including dates, identities of programmers, and copies of prior versions of the software. Given the importance of this electronic evidence to proving your case, you should consult with your client at the earliest possible opportunity about preserving and assembling its own electronic evidence, send timely and detailed preservation letters to the other side, and carefully craft discovery requests to obtain that key information.

In a software case, like any copyright case, the plaintiff bears the burden of proving

(1) valid ownership of a copyright, and

(2) infringement of that copyright. Given the complexities of many software copyright cases, especially if ownership is hotly contested, it often makes sense to bifurcate these issues, with ownership adjudicated first, then infringement and damages.

For plaintiffs, bifurcation puts the focus on the issue that often has the most emotional importance to them: confirmation of ownership of the software. The downside is that bifurcation tends to increase the length of the case, although success in phase one often leads to a favorable settlement of all or most aspects of the case. For defendants, bifurcation affords a means to defer, and potentially avoid, discovery into their own business activities, finances, customers, and licensees. Bifurcation also prolongs the case, an effect that often benefits the defendant. Some downsides are that if the defendant loses on ownership, the settlement value of the case is likely to increase significantly, the verdict could prove devastating in the marketplace long before the case reaches final judgment, and the loss could embolden the plaintiff to assert additional claims such as claims for contributory or vicarious infringement. For the court, bifurcation allows for a simpler and shorter trial in phase one. If the defendants prevail on ownership, then phase two (at least on the plaintiff’s copyright claims) is moot.And if the plaintiff prevails in phase one, chances are there will be a settlement rather than a phase two trial. Because of the technical nature of a software copyright case, early retention and use of experts is essential. If your litigation team does not include lawyers who themselves are conversant in software (and even if it does), you would be well advised to consult a software expert about what kind of information you should be seeking in discovery, and in what form. Examination of computer hardware used in the creation and modification of the software will also require you to engage a computer forensics expert with the necessary technical skills. Once the information begins to roll in, someone with expertise in computer software can help you to analyze what you have and what it

All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. means. Retention of an expert may also be necessary to adequately investigate whether your client has the claims (or defenses) necessary to proceed with the dispute. Use of non-testifying consultants is also recommended.Given most lawyers’ lack of sophistication regarding computer software, a consulting expert is a particularly valuable source of advice on what to obtain and analyze in discovery and may be more involved in the litigation planning process than you generally want a testifying expert to be, given the effect of the expert disclosure rules on materials that would otherwise be subject to work-product protections. A consulting expert can also be useful to conduct analysis from which your testifying expert should be insulated, such as a defendant’s initial analysis to determine the extent of potential liability for copying.The testifying expert can repeat favorable analysis conducted by the consulting expert and can conduct other useful analysis (such as performing the inspection of the opposing party’s computer). You should not discount your client’s own expertise in this regard. Particularly where you represent a programmer or a company with on-staff programmers, those individuals can provide useful insight into the process of writing computer software and likely sources of useful information. And if not precluded by a confidentiality order, they may also be able to review software produced in the case and provide helpful analysis of it. Expert copyright analysis of the parties’ software generally requires computer comparisons of the parties’ respective source codes to determine areas of potential similarity. Many different algorithms and programs exist for performing this analysis. The most common and basic type of comparison is to search for verbatim matches of portions of text between the parties’ source codes. There are many programs that can automatically compare the parties’source codes for verbatim matches.This type of comparison works well if there has been verbatim copying. However, given the ease with which an infringer can try to cover his tracks by changing variable names and programmers’ comments, or even converting the code to a different programming language, an analysis limited to verbatim copying will likely miss much infringement. Consequently, verbatim searches should be combined with other types of copy-detection comparisons for a complete analysis, such as a “structure matching” search that analyzes the programming structures in the parties’ software. Although some common structure is likely between different software simply as a result of commonly used programming practices, significant amounts of matching structure indicate copying. Computer software is often voluminous, and copy-detection search results are therefore often voluminous as well. And the search results usually contain false positives, including reports of matches that are explainable by such things as the requirements of a programming language or standard programming methods. Consequently, the expert must personally examine the parties’ software, usually by using the search results as an index of sorts to focus the expert on the portions of the software to examine personally. Although time consuming and expensive, this personal examination is indispensable to buttress the expert’s conclusions if the case goes to trial. And of course, when it is time to present evidence or argument to the court, having the benefit of independent or client experts will assist in translating the technical aspects of the case into a format that is user-friendly for people without a technical background. This will be essential at all stages of the case. Judges quite rightly rely on counsel to educate them on the issues and governing law. Unless the judge to whom your case is assigned has a strong technical background, you will also need to be prepared and able to explain the technical issues associated with all of your requests for relief. For example, if you are embroiled in a discovery dispute with opposing counsel who has produced the software as object code or in a format different from the source code’s native programming language, your ability to compel production of the source code will depend on your ability to explain to the judge why source code is necessary for a complete analysis of the software. By the time the case goes to trial, you have become technically savvy regarding the parties’ software and the primary technological issues at stake. The problem you now face is that the judge and jury will understand little of what you and the witnesses are prepared to talk about. Simplifying your presentation to make sense of the issues for the judge and jury can be far more difficult than you might expect. When presenting evidence and argument regarding the merits of the case, you must put forth your position in a manner that is accessible to those who do not themselves write or understand computer software. One strategy that is particularly useful is to remember that computer software, no matter how complicated or sophisticated its functionality, is just a set of instructions written in a foreign language.Those two concepts—instructions and foreign language—are well within our everyday experience and have the effect of demystifying the technical nature of the software. Everyone—lawyers, judge, and jury—has tried to give instructions, whether teaching a child to tie his shoes, writing out a recipe, or giving travel directions.And everyone is at least familiar with the idea that there are other languages in which those instructions could be expressed, even if they do not speak those other languages.You do not have to understand what the instructions are or be able to write them yourself in order to comprehend that one set of instructions is substantially similar to another. Preparing witnesses, developing demonstrative exhibits, and organizing your case with such commonsense analogies in mind will enhance your ability to convey the facts to the judge or jury. Also, do not underestimate the persuasive power of technological competence. Not only do you have an ethical duty to attain (or associate) the degree of skill necessary to comprehend and present the evidence in your case, but the lawyer who can work directly with the technological evidence gains credibility with the court and jury. Approaching the witness, obtaining from him a disc of software in evidence, and having that witness testify about what you display on the courtroom computer screen when you put the disc in your laptop is far more compelling than simply asking the witness to testify about software from a paper printout (or worse, no evidence at all). If you are comfortable with the necessary technology, you get better evi-

Verbatim searches should be combined with other types of copy-detection comparisons.

All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. dence from your own witnesses and in cross-examination, and your comfort with the technical subject matter goes a long way to establish your credibility with the judge and jury and makes them more comfortable in deciding the issues. Do not forget the importance of themes and narratives in the presentation of your case. Just because the subject matter is highly technical and an expert is almost certainly required does notmean that youmust depart fromaccessible, common themes in presenting your case. Copyright is still a very human story— someone who worked hard to create something unique should be able to usewhat shemade and be proud of it.Themes of theft, deception, and dishonesty are common in copyright cases and make even highly technical subject matter more accessible to the judge and jury.

The Copyright Act allows the court, once liability is determined, to “award a reasonable attorney’s fee to the prevailing party as part of the costs.” 17 U.S.C. § 505. Fees are awardable to the prevailing party, whether it is plaintiff or defendant. Whether to award fees is left to the court’s discretion. The factors a courtmay considerwhen exercising its discretion include, but are not limited to, “frivolousness, motivation, objective unreasonableness (both in the factual and in the legal components of the case) and the need in particular circumstances to advance considerations of compensation and deterrence.” Fogerty v. Fantasy, Inc., 510 U.S. 517, 534 n.19 (1994). Courts even allow fee awards to include costs and fees associated with the prosecution or defense of other claims, such as unfair competition and defamation, if those claims and the copyright claims involve a common core of facts or are based on related legal theories, and can include fees incurred in connection with interlocutory appeals. E.g., Twentieth Century Fox FilmCorp. v. Entertainment Distributing, 429 F.3d 869, 884 (9th Cir. 2005), cert. denied, 126 S. Ct. 2932 (2006). In a sense, the key lesson of our experience—and this article— is that litigating the software copyright case is just like litigating any other case.You will need to investigate the claims, assemble the evidence, and set forth your proof in a compelling, comprehensible manner. The fact that a case involves software just requires you to take the necessary steps to plan for and understand technical subjectmatter, just as you would in a case involving technical evidence of any other kind, be it medical, scientific, or statistical. The peculiar nature of software, particularly its manifestation in both object and subject code, simply dictates some of the ways in which you will take the steps on that familiar path.

Click here to view a PDF of original article

Related Industries