Risk management today has its main roots in decision theory paradigm (Friedman and Savage 1948). It consists in making the optimal choice between given possible decisions and probable states of nature. In this paper we extend this model to include a design capacity to deal with risk situations. A design perspective leads to add a new action possibility in the model: to design a new alternative to deal with the probable states of nature. The new alternative design might also "create" new risks, so that a design perspective leads also to model the emergence of new risks as an exogenous "design process". Hence a design perspective raises two issues: can we design an alternative that would lower the risk? Does this new alternative create new risks? We show (1) that minimizing known risks consists in designing an alternative whose success is independent from all the known risks – this alternative can be considered as a generic technology. We show (2)that the design of this generic technology depends on the structure of the unknown, ie the structure of the space generated by the concept of risk-free alternative. (3) We identify new strategies to deal with risks as dealing with the unknown.