Abstract

The catastrophic risks posed by new technologies such as killer robots and geoengineering have triggered calls for halting new research. Arguments for restricting research typically have a slippery‐slope structure: Researching A will lead to deployment; we have decisive moral reasons against deployment; therefore, we should not research A. However, scientific uncertainty makes it difficult to prove or disprove the conclusion of slippery‐slope arguments. This article accepts this indeterminacy and asks whether and when it would be permissible to restrict research under empirical and normative uncertainty. The argument starts from the ethical framework for the regulation of scientific research with human subjects and offers modifications to adapt it to the purpose of restricting new technologies. Two main questions arise in the process: whether it is permissible to impose restrictions at the research stage to prevent harms that will arise from the use of a technology and whether it is permissible to restrict research preemptively on the grounds of public fear and anxiety, before there is sufficient evidence establishing the risk of harm. I answer both questions in the affirmative and defend this position against objections.