Steve Brenner, assistant professor and leader of a computational genomics research group at the University of California, Berkeley, said he fears that many academic bioinformaticists are unaware of a legal risk they face on a daily basis: contributing to open source software projects without explicit permission from their institutions.
While many employers have clauses in their employment contracts that restrict the creation and use of open source software, bioinformatics programmers at universities are often not as attuned to copyright issues as their industry counterparts. This fact, Brenner said, raises the possibility that a good portion of biological open source software is currently being produced illegally.
Brenner — a founding coordinator of the bioperl open source effort — was alerted to this issue when he was first offered his position at Berkeley. Upon realizing that the university’s default software license was incompatible with his continued work on open source, Brenner requested a formal variance from the normal license as a stipulation of his employment contract.
Brenner did succeed in renegotiating his employment contract with Berkeley, but the experience turned out to be a painstaking one — far longer, much more complicated, and a great deal more expensive than Brenner had anticipated, largely because his request was unprecedented within his department.
“Generally everyone at the University was very helpful and tried to deal with it rapidly,” Brenner said, “but since none of them had any experience with requests like this, it ended up taking much longer than it might have if it were a regular occurrence.”
But while Brenner’s request may have been unusual, his circumstances are certainly not. Realizing that a large number of his colleagues and peers in the bioinformatics community may be facing a similar risk, he and other members of the Open Bioinformatics Foundation have set out to raise awareness of this issue. Open-bio is hosting a discussion list on the topic (http://open-bio.org/mailman/listinfo/authors) and is looking into drafting a standardized contract that open source developers could use in order to ensure legal protection for their work.
“The hope is that many people will want to solve these sorts of problems and therefore the community could come together to develop a contract addendum similar to the contract that I have with Berkeley,” Brenner said.
After speaking about his experiences at a talk at the Bioinformatics Open Source Conference in Copenhagen July 19, Brenner said there was a good deal of interest from people who didn’t realize they were currently violating their institutions’ policies.
“People in industry tend to be more aware that they can’t produce open source without getting permission from their company, but academics may not carefully consider intellectual property ownership. In many cases, the university does own the rights to their work,” he said.
This misconception is fairly widespread, according to Bradley Kuhn, vice president of the Free Software Foundation, an advocacy group that supports the continued development of free software, which it stresses as distinct from open source. “Many people in universities typically assume that they’re working in a free environment where their work is going to be made available to the public in a useful way,” Kuhn said. “These days it’s up to every researcher, faculty member, graduate student, and even undergraduate student at a university to find out what the policy about copyrights and patents is and find out if the copyright to their work is going to be held by the university.”
Larry Rosen, an attorney who consults for the Open Source Initiative, a non-profit corporation that supports open source development, agreed that university employees often don’t consider their work in terms of a traditional employer-employee relationship — and that this situation puts them in danger of violating their organization’s policies.
Rosen stressed that under copyright law, “anything that you do is owned by your employer and you’re not free to give it away … As a general matter for any employee, if you want to contribute to any open source project, whether it’s your own thing or you contribute to an open source project for someone else, you have to make sure it doesn’t get treated as a work-for-hire. That applies to any employee-employer relationship, whether it’s a university or anyone else.”
The issue seems to be coming to a head in the academic world now, as more universities are exploiting the revenue stream made possible by their copyright and patent holdings. “If you’re a software developer, the university holds rights to your software, but if you’re an English professor or Law professor and publish a book, they’re not the least bit interested in copyright,” said Thomas Field, an attorney at the Franklin Pierce Law Center affiliated with the Association of University Technology Managers. “It’s hard to justify having any stake in poetry, but some of this software is worth serious money. When money starts to appear, people want to have a stake in it.”
With the exception of the University of Texas system, which amended its copyright policy in 1995 to support the General Public License, most universities prohibit open source for a number of reasons beyond the potential loss of revenue it represents. Such policies are in place to ensure appropriate acknowledgement, prevent theft, and avoid liability, for example.
Veronica Lanier, a senior licensing official in the Berkeley office of technology licensing, said the university’s copyright policy “provides that the university owns the copyrightable material at the time of creation, but it also provides that authors of the copyrightable work will share in license revenue.”
Because there is no license revenue from an open source project, Lanier said, every member of a research group who may have contributed to open source software would have to “acknowledge that they are giving up certain rights that they would have to receive a share of royalties.”
“Universities are acting responsibly in trying to protect their intellectual property,” Brenner noted. “There are a lot of good reasons why institutions would want to retain copyright. What we’re trying to do is convince them that our scientific needs are compelling and that we think these needs are more important than the reasons why they have restrictive copyright controls.”
Brenner’s renegotiated contract permits him and members of his research group to produce open source software with three requirements: the funding body has to agree, everyone who has an involvement in producing it has to agree, and it can’t violate any laws.
Lanier said that Brenner’s foresight in requesting the waiver before beginning employment at the university actually made the process much easier than if he had waited until he had begun to work there. However, noting that even this “fast track” situation required an unreasonable amount of time, effort, and legal expense, Brenner said a standardized employment contract would be an advantage both to producers of open source software and to their employers.
“The advantage of having uniform contracts would be that [employers] would have to learn about the contract only once. From then on, they could agree to it on a routine basis without having to invest their energies and legal efforts into investigating it,” Brenner said.
The effort is still in a nascent stage, Brenner said, and the Open Bioinformatics Foundation is currently exploring the logistics of drafting a standardized open source author’s contract as well as its funding options for supporting legal representation. “It will probably involve also getting institutions involved to make sure that the contract is generally something they find acceptable,” Brenner said.
The main issue, Brenner maintained, is that “right now people are violating their institutions’ policies. Usually that’s because they have never asked the institutions to change rather than because the institutions have refused to change.”