Synthetic biology is the attempt to design novel biological devices to improve life, such as bacteria that can produce energy or drugs cheaply or new biological therapies. But the field also has a potential dark side.

Scientists can order expressly designed chunks of DNA from a number of DNA synthesis companies around the world and then fuse these bits together to create new biological "parts." And the same technology that could lead to valuable inventions can also be used to make deadly bioweapons -- hypothetically, terrorists could order DNA to recreate the smallpox virus or design an even more deadly pathogen. While most experts doubt that fringe groups currently have the ability to pull off such a feat of biological engineering (see "The Knowledge"), scientists worry that, as DNA synthesis technologies quickly become cheaper and more accessible, the possibility of nefarious use will grow.

At the recent Synthetic Biology 2.0 (SB2) conference, held in Berkeley, CA, scientists devoted an entire day to a single topic: the security concerns of the fledgling field. The result was an evolving community declaration outlining how scientists and companies should behave to ensure both openness and security of research.

The researchers pledged, for example, to develop better software to detect when orders for dangerous DNA sequences have been placed with DNA synthesis companies, and they recommended that scientists work only with companies who use such software. Drew Endy, a biological engineer at MIT and one of the organizers of the conference, told Technology Review about these discussions on security and what he hopes to accomplish with the new declarations.

Technology Review: What issues are you most concerned about?

Drew Endy: As we've made progress in synthetic biology research during the past few years, we've also recognized almost immediately that there are a number of issues at the interface of the technology and society. Briefly, these issues can be organized into four topics: safety and security, ownership and sharing, understanding and perception, and community organization. It is irresponsible to develop any technology without also directly addressing the associated nontechnical issues. For example, today, not all DNA synthesis companies check what they make. This was demonstrated recently by the Guardian, which published a front-page article stating you could order a piece of smallpox DNA through the mail. We tried to dissuade the journalist from going forward with this stunt.

TR: So the community put together some explicit declarations, sort of a set of goals and guidelines, to try to begin to address security practices around DNA synthesis. What are the main points outlined in the declaration?

DE: First, we want to make sure that the use of DNA synthesis technology is subject to the same community and institutional oversight mechanisms that have been used successfully with recombinant DNA work for the past 30 years. The main challenge here is that DNA synthesis technology is becoming easy to access anonymously via the Internet. Thus, we are asking DNA synthesis companies to work together to develop an open framework that can be used to ensure that all synthesis orders are placed by qualified individuals who have proper authority for handling the requested DNA. Because such a framework needs to be used everywhere in order to be effective, we are encouraging individual researchers to patronize DNA synthesis providers that employ the framework. The remainder of the declaration is a call to continue open and constructive dialogs that directly address the remaining topics, for example, ownership and sharing.

TR: Are you calling for an all-out boycott of companies that don't check what DNA they are synthesizing?

DE: Not yet. The technology for checking orders isn't perfect and would not be practical for use in some settings, for example, high-volume synthesis of very short DNA fragments. The goal for now is to strongly encourage otherwise competing companies to work together and to coordinate with government and others to solve the problem. As improved methods are developed, implemented, and deployed, we expect that adoption of best practice will be widespread. If it turns out that no progress is made, stronger actions may be warranted.