Abstract

It is sometimes suggested that new research in such areas as artificial intelligence, nanotechnology and genetic engineering should be halted or otherwise restricted because of concerns about possible catastrophic scenarios. Proponents of such restrictions typically invoke the precautionary principle, understood as a tool of policy formulation, as part of their case. Here I examine the application of the precautionary principle to possible catastrophic scenarios. I argue, along with Sunstein (Risk and Reason: Safety, Law and the Environment. Cambridge University Press, Cambridge, 2002) and Manson (Environmental Ethics, 24: 263–274, 2002), that variants of the precautionary principle that appear strong enough to support significant restrictions on future technologies actually lead to contradictory policy recommendations. Weaker versions of the precautionary principle, which do not have this feature, do not appear strong enough to support restrictions on future technologies.