This article proposes to link normative issues generated by artificial intelligence in the legal sphere to the more fundamental problem of "institutional design" by Lon L. Fuller. The optimization of state resources promised by the deployment of artificial intelligence will take place in return of a more preventive and unilateral enforcement of the law, in other words a managerial, "noncontentious" enforcement of the law. By splitting the concept of state governance into its "legal" and "economic" ramifications, Fuller intended to combine these ramifications in a project aspiring to good order and workable social arrangements. A reactualization of Fuller's writings will allow us to situate the scope of artificial intelligence within this classification, and will thus shed light on the normative issues linked to the use of artificial intelligence by the state.