It is a little weird to think of engineered bacteria living in your mouth or your gut, fighting cavities or Crohn's disease. I'll admit I feel a twinge just thinking about it. But is that because I have some intuition of the risks of ingesting such creatures? I doubt it. I think it's just focusing my attention on the prospect of some living thing living inside me. But we're already packed with thousands of species, and we regularly get infected (or maybe I should just say colonized) with new microbes. We even purposefully take in bacteria for our well-being when we proudly spoon yogurt into our mouths.
Does that mean that there's absolutely no possible risk from swallowing bacteria loaded with human immune signal genes? No. We might guess that these genes would put these bacteria at a competitive disadvantage against the other bacteria struggling to survive in our guts. But that's a hypothesis. Judging the risks of these kinds of organisms is a lot harder than judging the risks of a cigarette or an asbestos factory. The microbes can evolve and they can pass their genes on to other microbes. Their effects may depend on the other species around them. This is true both for engineered microbes inside of us and for engineered crops and other free-living creatures in the wild. In both cases, we're introducing new players into ecosystems, and ecosystems are horrendously complex things. I've been looking over a new report from the National Academies of Sciences on the risks of genetically engineered organisms, and it deals mostly in experiments that still need to be done, not insights from past experiments. Are we really just getting started thinking about this stuff?
People have been thinking about the applications for a long time, but the risk:benefit calculation has been on hold. It may be that there is no feasible way to estimate the risk.
I was reminded of something I read earlier today, about the possibility of using genetically-modified viruses to treat cocaine addiction:
(I think the first page is open-access, and it is only a one-page article)
The only way to test it for safety is to try it in humans. Once you do that, the genie is out of the bottle. While the risks are almost certainly very small or nonexistent, the technology is so new that certainty will be highly elusive.
If there is no feasible way to assess the risk, what do you do? In that case, what is needed is a policy decision about what to do when there is a promising technology, but no way to assess the risk. This would be policy that is guided -- but not determined -- by science.