KERALA RISING · FROM EXPORTING PEOPLE TO EXPORTING IDEAS contract, is built for one use case, and accumulates no institutional capability for the state. Seven civic outcomes procured this way mean seven incompatible silos, seven times the cost for the shared capabilities, and nothing the state owns at the end. The framework’s institution-and-reuse model is the structural answer: build once, own forever, configure cheaply. The vendor model is not merely more expensive; it cannot produce the compounding that makes the marginal program cheap. Alternative four: a startup or private platform The fourth alternative is to let a startup build a civicengagement platform and run it commercially. This brings speed and product discipline, which are real advantages. But it fails the trust test that civic AI most needs to pass: a private platform with equity holders has incentives to monetise data, to favour paying interests, and to be acquired or shut down. Citizens are right to be warier of a private company recording their civic actions than of a publiclyowned, publicly-audited institution. And the moment the platform is sold, the public interest it was meant to serve is subordinate to the acquirer’s. The framework by design forgoes the startup’s speed for the institution’s permanence and trust, because in civic infrastructure those are worth more. 304