As a computer science student in the '80s, I was well-steeped in the pioneering spirit that led to the modern open source code movement. Back then, we were far too busy exploring the potential of digital computers (remember artificial intelligence?) to concern ourselves with notions of ownership or commercialization. Those who were lucky enough to break new ground were happy to share their accomplishments with the programming community in the hopes of earning some modest peer recognition while helping others take the next step.

When computer programs became widely integrated into commercial society in the '90s, the free software movement began looking for ways to formalize the traditional openness among programmers. This formalization took the form of open source code licenses. These licenses allowed programmers to freely use "open source" software code in commercial products if they agreed to certain open source terms. Unfortunately, there was little uniformity between open source licenses, and "acceptable" open source terms ranged from the innocuous (using proper copyright notice) to the egregious (agreeing to license any of your own programs that contain open source code for free). While sophisticated companies certainly recognized the cost benefit of using "off the shelf" software, the attendant legal complexities have not always been routinely considered.

The recent Goldman Sachs trade secret trial provides yet another reason to carefully consider the use of open source software. In early December, Sergey Aleynikov was convicted of stealing trade secret computer code he developed for his former employer, Goldman Sachs. According to court documents, the government alleged that Aleynikov took the software code to enable him to develop a competing program for his new employer, Teza Technology.

In developing proprietary code for Goldman, however, Aleynikov apparently incorporated some open source elements. In defending against the misappropriation charges, Aleynikov's lawyers argued that he did not intend to take Goldman trade secret code, but, rather, he was merely trying to take the open source code he was entitled to take under the applicable open source license. This defense added a second level of complexity to what ordinarily would have been an open-and-shut case. Not only would a jury need to decide that Aleynikov took software with him, they would also need to determine that the purloined software was open source. Given the vagaries of many open source licenses, this second question is not always easy to answer.

While it appears that the government was able to overcome the "open source" defense in the Goldman case, other companies may not be so lucky. To prevent "open source" issues from overriding proprietary protection, companies need to carefully consider the use of open source code in their proprietary systems.

If open source code is used, technical directors need to ensure that it is physically and logically segregated from proprietary source code elements. In-house counsel should also be aware of the origin of any open source elements being used so that they can evaluate the applicable open source license terms. Finally, companies should not let programmers make the decision on whether or not to include open source code in proprietary systems. The legal and business implications of open source are far too nuanced to allow decisions to be made real-time during software development.

Mark Scarsi is a partner in the Intellectual Property Group of Milbank, Tweed, Hadley & McCloy, LLP, resident in the Los Angeles office.

Read Mark Scarsi's previous column. Read Mark Scarsi's next column.