The widespread use of data to represent reality and the development of artificial intelligence (AI) are reshaping the modes of governance in contemporary societies based on social contract theory, replacing traditional political mediation (representation, controversy, accountability) with optimization and prediction. This thesis focuses on social contract theories as the foundation of contemporary democracies and addresses two questions: what is the political contract becoming in the digital age, and how can it be rebuilt without renouncing popular sovereignty? The challenge is to show that data, metrics, dashboards, and algorithms are disrupting public action and redistributing authority among decision-makers, experts, operators, and platforms. In doing so, the contemporary evolution of digital technologies is eroding the political, democratic, and emancipatory promises of contractualist theories (notably Rousseau's "general will," as the basis for popular sovereignty) and establishing deterritorialized private sovereignties and private institutions in cyberspace that are inherently illegitimate because they do not emerge from the social contract. We therefore develop a twofold diagnosis: on the one hand, the contract is being emptied of its functions by the outsourcing of judgments in favor of automated and rationalized procedures; on the other hand, historically, the ideology of "governance by numbers," now digital governance, prolongs old dreams of rationalization that tend to neutralize dissent and deliberation. Consequently, beyond ethics centered on compliance and usage, this thesis argues for the need to develop a political and institutional approach that makes explicit the values incorporated into systems and restores to the community, to the people of the contract, the ability to define its own ends. The thesis thus proposes four complementary approaches: (1) repoliticizing technology by making technological choices the subject of debate and decision-making; (2) rebuilding democratic mediation in the age of computation (transparency of models, safeguarding of controversial issues, accountability, participation) to ensure that digital technologies serve democratic objectives; (3) rebalancing power through institutional counterweights; (4) formally integrating data, AI and Big Tech corporations, into the contract, setting purposes, limits, and responsibilities within a common normative framework. The aim is neither to sanctify technology nor to decree the end of politics, but to draw up a renewed social contract in which technological developments around digital technology and AI remain under public control and serve collectively and democratically determined ends.