The Ownership
of Ideas in Computer Software:
A Contrast Between Trade Secrets, Copyrights, and
Chip-Mask Protections
John W. Snapper
IV. The Methods of Discovery in Reverse Engineering
The issue in Kewanee was whether the ideas behind an invention could be held as a trade secret during the patent application process and continue to be held as a trade secret if the application is denied. If there were a straightforward conflict between federal patent policy and state trade secrecy policy (one excluding and the other promoting the protection of ideas), then the federal policy might have “preempted” state law in this specific case. The preemption issue has also arisen for the copyright law in the context of computer software, and it will (we may assume) sooner or later also arise for chip registration. By comparing policy on how copyright and chip-registration, we can get a sense of the extent to which these forms of intellectual property extend to protections for ideas.
Prior to the use of copyrights for software, it is hard to imagine someone attempting to keep information secret while at the same time putting it into a copyrighted manuscript. The nature of the subject matter of copyrights precluded dual use of trade secrets and copyrights and thus precluded conflicts leading to disputes over preemption. This changed with copyrights on computer software. Computer software may be copyrighted while largely in machine executable form. It will then require considerable effort at reverse engineering to discover the algorithms contained in that functioning software. In fact, the copyright could be used to prevent discovery of secret information encoded in copyrighted software because the process of reverse engineering might involve steps of copying that infringe the copyright. I personally consider this an abomination of the copyright tradition (see Snapper, 5 Social Epistemology 78, l991). But I would not wish to predict the direction that the law will take in this area.
Dual use of copyrights and trade secrecy is presently much debated in the law reviews, as a form of the issue of “preemption.” It is entirely possible that we may see a new version of the argument used in Kewanee that justifies dual usage on the grounds that copyrights and trade secrets protect different things (the expression of ideas and ideas themselves). (This philosophically displeasing argument was used in Warrington Associates v Real-Time Engineering Systems (522 F. Supp. 369), although it does not have the weight in copyright law that it holds in patent law.) The dual usage of trade secrets and copyrights contemplated here is, in fact, much more extreme than that upheld in Kewanee. Kewanee considered the status of trade secrets in rejected patent claims. The copyright issue concerns the status of trade secrets in software with recognized copyrights. (This form of dual protection is clearly rejected by patent law. In White Consolidated v Vega Servo-Control the court refused to grant trade secret protection to software required for the operation of a patented machine, since patent law requires disclosure of all operational information; 713 F. 2d. 788 (1983).)
If developing copyright policy (a) recognizes dual protection of software as a trade secret and copyrighted code, and (b) further recognizes that copying involved in engineering may violate a copyright, then reverse engineering of copyrighted software will be severely limited. Some form of reverse engineering would still be possible. Software could be tested against certain conditions to see how it reacts, and inferences then drawn about its underlying algorithms. Duncan Davidson has called this “black box” reverse engineering, since the engineers never look internally into the code of the software. Davidson argues that reverse engineering of copyrighted software should be restricted to black box investigations, as not only the one form of reverse engineering that is clearly within the bounds of the law but also as a form of study that encourages R&D among competing firms (see 47 Univ. of Pittsburgh Law Review 1037 (1986)).
In contrast to the restrictions on reverse engineering that might be entailed by copyright policy, the SCPA #906 specifically and loudly endorses any form of reverse engineering for the discovery of all information held in a registered chip. Once again, this contrast with the copyright law is to a large extent the reason for the inclusion of that section. On the other hand, the fact that information has been placed into a registered chip need not exclude its continued protection as a trade secret for some period. By so strongly promoting reverse engineering, the SCPA tacitly admits that there may be secrets in a chip mask. I therefore conclude that the SCPA clearly does not preempt simultaneous use of trade secrets, such as is debated in copyright law. Imagine for instance, that a chip is registered and available for reverse engineering. But a competitor, rather than spend the year needed to reverse engineer, learns the information from an employee of the chip owner. Could the chip owner sue for lost profits for the one-year period, claiming misappropriation of a secret? (The analogous situation in copyright law would be complicated if copyrights are found to preempt trade. A suit for unfair trade practice might still be possible, although it is unlikely to be viewed as misappropriation.)
The SCPA makes a point of spelling out the relation between its protection and trade secrets differently than in developing copyright policy. It is conceivable that Congress’s explicit discussion of these points in the SCPA will be seen by the courts as justification for a looser interpretation of the copyright preemption issue. The argument is that since Congress has decided not to address the issue for copyrights like it did for chips, Congress must have intended a different standard. (I hope we do not, in fact, see this bad argument made by the courts.) But in any event, the differences between copyright and chip law illustrate the wide range of give and take in what we interpret as a conflict between federally mandated intellectual property and the ownership of ideas.
Illinois Institute of Technology
Go to: Why Software Should Be Free – Stallman
Home > Research Resources > Software Ownership & Intellectual Property Rights > The Ownership of Ideas in Computing Software
HOME |
IN THE NEWS | RESEARCH RESOURCES
TEACHING
RESOURCES | STUDENT RESOURCES | LINKS
The Research Center on Computing & Society
at Southern Connecticut
State University
501 Crescent Street • New Haven, CT 06515
Director: (203) 392-6790 •
e-mail: webmaster@computerethics.org
© 2000 – 2007 – Research Center on Computing & Society