The history of American law is rich with examples of large companies pursuing small competitors in the courts, there to find not only defeat in the case at hand, but to discover that their actions forced a change in the law with implications that go far beyond the immediate dispute. Computer Associates International, Inc. v. Altai, Inc. is such an example. Computer Associates’ single-minded pursuit of Altai not only resulted in an expensive defeat, but may also have greatly narrowed the scope of copyright protection for all developers in the software industry.
The case began inauspiciously. Altai hired a programmer from CAI. Unbeknownst to Altai, its new employee brought CAI source code with him and used it to develop an Altai program. CAI discovered its former employee’s actions and promptly brought suit against Altai. Altai then conducted an internal investigation and learned that a program developed by the former CAI employee consisted of 30% CAI code. As a result, Altai developed a second version of the program, which was written by programmers who had never seen either the misappropriated CAI software or the software written by the former CAI employee that included the CAI code.
In the litigation, Altai conceded that it was liable for copyright infringement for the first version of the program. However, it claimed that the second version did not infringe CAI copyrights because the software had been developed under “clean room” conditions and contained no CAI code whatsoever. CAI maintained that the second version copied the structure of CAI’s software, and therefore was still infringing.
The trial court awarded CAI $364,000 in damages for the admittedly infringing program. However, the trial judge held that Altai’s clean room version of the program did not infringe CAI copyrights. As we noted in our January, 1992 issue, the trial judge rejected the 1987 landmark holding inWhelan v. Jaslow , which held that copyright law protected the “structure, sequence and organization” of a program. The trial judge also held that the “behavior” of a program cannot be the subject of copyright protection, since behavior is akin to a “process,” “system” or “method of operation,” all of which are excluded from copyright protection.
Had CAI accepted its partial loss in the trial court, this decision would almost certainly have faded into obscurity. A decision from a single trial court, in an area of law where there are many ongoing cases, rarely has a profound effect on the law. It was only the fact that CAI had appealed the ruling to the Second Circuit Appeals Court in New York City — long recognized as the most important federal circuit in the area of copyright law — which led to our scrutiny of the case for our readers earlier this year.
The Second Circuit’s opinion, issued on June 22nd, shocked the industry. It criticized Whelan ‘s standard for judging copyright protection for software as being overbroad. It held that Whelan’s use of the terms “structure, sequence and organization” demonstrated a flawed understanding of the method of operation of a computer program. Finally, it attributed Whelan’sflawed reasoning to the Whelan court’s “outdated appreciation of computer software.”
The Second Circuit went on to posit a new, three-part test. First, the allegedly infringed program must be broken down into its constituent structural parts. Second, the program must be stripped of all elements which are dictated by efficiencies, external factors (such as compatibility requirements, design standards and accepted programming practices) and public domain elements. A “core” of protectible expression may remain. In terms of a work’s copyright value, “this is the golden nugget.” This “nugget” must be compared against the allegedly infringing program (which is not subjected to the “stripping” process), to determine whether the defendant has copied any part of this protected expression.
Many of the opinions in the most important prior software copyright cases discussed the policy implications of their holdings and observed that broad copyright protection is necessary to encourage investment and creativity. The Computer Associates court came out very differently on this issue. The court rejected the argument that companies would not invest the extensive time, energy and funds required to design innovative software unless their programs received broad protection. The Second Circuit also expressly disagreed with the Whelan court’s holding that software copyright protection was necessary to provide the proper incentive for programmers by protecting their efforts. The court even expressed doubt as to whether copyright law was appropriate for the protection of software works, and suggested that Congress revisit this question.
TLB Comment. The Computer Associates decision is a lengthy, complex opinion, and its holdings and implications go far beyond what we can present here. However, a few observations are in order. First, the case is unclear on whether, under its view of copyright law, program structure retains any copyright protection at all. With masterful understatement, the Court of Appeals noted: “to be frank, the exact contours of copyright protection for program structure are not completely clear.” However, the court did nothing to clarify those contours, although it did not overrule the trial judge’s holding that a program’s “behavior” cannot be the subject of copyright protection. This bodes ill for the future of infringement suits based on claims that structure has been copied.
Second, we believe that the court intended its observations, and its method of analysis, to be applicable to cases involving user interfaces as well. There is no logical distinction for treating user interfaces differently, and many of the cases the court refers to involve claims of interface infringement, as opposed to code infringement.
Third, it is worth noting that the Court’s statement that copying would not constitute infringement if it were necessary to ensure “compatibility requirements of other programs with which a program is designed to operate” and “demands of the industry.” when Lotus sued Paperback Software in federal court in Massachusetts, Paperback unsuccessfully made both of these arguments. The questions thus arise as to whether these arguments would have succeeded had the case been brought in the Second Circuit, and what impact the CAI holding will have on the Lotus/Borland litigation which involves many of the same issues.
CAI now faces its most difficult decision: whether to appeal the ruling to the U.S. Supreme Court, or live with an unfavorable, but nevertheless limited, decision from the Second Circuit. Given the clash between the holdings in Whelan in the Third Circuit and Computer Associates in the Second Circuit, and the importance of the issues to the computer industry, we believe there is a good chance the Supreme Court would hear the case if CAI asked it to do so. At stake for the entire industry in CAI’s decision is the fact that a definitive ruling by the Supreme Court would establish the state of the law for the entire U.S. Until such a ruling comes from that court there remains room for the various federal courts to diverge or disagree.
Postscript: Other federal courts have, for the most part, followed the Altai test. The Altai case itself is still proceeding in the district court, where collateral issues relating to trade secret claims continue to be argued. Thus, there has been, as yet, no opportunity for appeal to the Supreme Court.