Intelligent Speed Adaptation (ISA) could improve road traffic safety significantly. Large scale implementation of ISA, however, is hampered by uncertainties, such as the level of user acceptance, reliability of the technology, unexpected costs, etc. In this paper we suggest an adaptive policymaking approach that would enable ISA implementation to proceed despite these uncertainties. We propose five steps: (i) specifying the policy problem, (ii) assembling a basic policy, (iii) specifying the rest of the policy, (iv) learning from real world experience with the policy, and (v) changing the implemented policy. The adaptive approach is supported by exploratory analysis. Using a simple road safety model, the analysis involves making computational experiments across multiple plausible states of the system and multiple scenarios. We demonstrate how to use the insights from the exploratory analysis to design a basic policy and how to gain knowledge over time about the behavior of the system in order to resolve some of the uncertainties. By monitoring the change in mean speed to detect changes in the level of acceptance, the policy can be adapted to unfolding events. For example, for two target groups of drivers, young and middleaged ones, a policy that targets more young drivers with mandatory ISA is shown to be robust.