Rapid climatic, natural and societal changes are altering the ways natural hazard risks are represented in societies, and in turn disrupting the ways people respond to these hazards. This poses an important challenge to how societies (re-)build institutions for governing or controlling risks. Institutions are systems of rules, norms and decision-making processes that structure our social interaction and practices. They organize how people define, plan for, and manage natural hazard risks; indeed, they create notions of risk. Going deeper, social sciences have defined institutions by the underlying “culture” on which they are built; the symbols, principles, core beliefs, and cognitive scripts that give institutions meaning. The culture structures how institutions represent the intertwined natural and social world that gives rise to natural hazard risks. Cultures work as a script for classing risks; giving people cues on how to understand and interpret the dangerous situations they find themselves in. Modern institutions are increasingly shaped by techno-scientific cultures, defining hazards and risks by their technically framed probability of physical harm, often expressed in terms of loss and damage. This risk quantification, and aspirations for precision, can give a false sense of control. But climatic change is already undermining, and threatening to undo, many of the long-held representations of natural and social order (and risk to this order) that steer institutions. Current case study research, in different places around the world, shows how climatic change is altering the way institutions interpret the natural hazards they manage in Bangladesh, New Zealand, and Norway for example. Dramatic climate change is confounding institutions’ cultures of risk quantification, and protection, shaking their claims to control natural hazards and undermining public trust in these institutions. One response is that institutions change the ways they define and class hazards, so that ordinary hazards are amplified as extraordinary. Faced with risks that are going beyond their experience and control, some institutions are compelled to unreflexively amplify well-intentioned protection-based responses, with at times unforeseen and disastrous consequences. Cases in Bangladesh and Norway both show how rushed river engineering works can evoke resistance from local communities. Emergency coastal protection can also have deleterious long-term social-ecological impacts, as experience shows in New Zealand. Scholars and practitioners alike recognize the need for critical reflection on how institutional cultures alter natural hazard risks according to climatic and other changes. This reflection is practical work that affects how people operate in institutions every day. It is structural work, as institutions change their rules as they learn more about risks. And it is work of social change, with social groups inside and outside institutions increasingly vocal in their criticism of changing climate risk framings. Case studies illustrate processes of institutional change, but equally, the resistance of institutions to change their cultures and notions of risk.