The idea of randomness is mostly excluded from international legal argumentation. If we need law at all, we are told, it is precisely to avoid arbitrary fortuity. Nonetheless, the exclusion of randomness renders international law structurally incapable of dealing with general risk issues, be they external or manufactured. The core of the problem is the notion of causation. International law seems to be infused by a model of causation that excludes any consideration of randomness. The law of state responsibility and certain elements of international trade law bear witness to this point. Randomness, however, is Janus-faced, and risk is its correlate aspect. By excluding randomness, risk is also left out. Therefore the model of causation embedded in international legal language makes that very language incapable of framing the ideas that a risk society needs to express. It is not surprising, then, that risk societies turn to other languages to express their needs, as is evidenced by the WTO SPS disputes. In this context, international law seems to become a broker of expertise, which refers to the relevant epistemic community that is needed to ‘get the job done’, and abandons any aspiration to holding an independent normative pull in itself.