Abstract
High-stakes (dangerous, catastrophic) risks take on a wider profile as progress unfolds. What are the impacts of technological and social change on the risk landscape? Due to the complexities and dynamics involved, we can only answer these questions approximately. By using the concept of fuzziness, we can formalize our imprecision about high-stakes risk, and therefore place their management on a stronger footing. We review here the impacts of fuzziness, i.e., knowledge imperfection, on high-stakes risk management, including its implementation via computationally intelligent decision aids.