Algorithmic social contract

The algorithmic social contract is the implicit agreement between individuals and the algorithmic institutions that govern them — a set of mutual expectations about what data will be collected, how it will be used, and what recourse individuals have when those expectations are violated. Unlike the classical social contract, which is theorized as a relationship between citizens and the state, the algorithmic social contract is a relationship between users and platforms, between patients and clinical decision support systems, between drivers and navigation algorithms. It is rarely written, never negotiated, and universally asymmetric: the algorithmic institution knows the user in granular detail, while the user knows almost nothing about the institution.

The algorithmic social contract is broken by design. The institutions that benefit from algorithmic governance have no incentive to make their operations transparent, accountable, or contestable. The contract is not broken because users are naive or because regulators are incompetent; it is broken because the feedback topology of algorithmic institutions rewards opacity and penalizes transparency. Any meaningful algorithmic social contract would require structural interventions: mandatory disclosure of decision criteria, user-controlled privacy budgets, and algorithmic circuit breakers that interrupt positive feedback loops before they produce harm. These are not technical fixes but constitutional requirements: they define what it means for an algorithmic institution to be legitimate.