AI as a Civilisation-Scale Audit

This updated and expanded essay argues that artificial intelligence functions as a civilisation scale audit, exposing structural weaknesses that long preceded the present AI boom. The original argument centred on the glide, a pattern in which real harms pass through AI and become misread as problems of AI alone, while the older systems that made those harms possible remain untouched. The expanded version keeps that core argument, then widens the audit across additional domains, including insurance, welfare administration, cognitive development, non Western knowledge systems, grief technology, and autonomous weapons. The essay traces AI’s arrival inside systems already shaped by surveillance capitalism, insecure work, platform monopoly, fragile public institutions, extractive infrastructure, and concentrated corporate power. It then treats AI risk reports, safety failures, automated decision systems, and institutional responses as diagnostic evidence, showing where legal frameworks, welfare systems, educational models, labour protections, public accountability, and moral language no longer match the scale of the systems they are expected to govern. The updated version adds five major areas to the original analysis. It examines insurance as a form of private algorithmic governance, where risk pricing increasingly shapes access to care, housing, mobility, and security. It expands the welfare discussion into an account of algorithmic cruelty at scale, where automated systems intensify punitive policy through speed, opacity, and administrative distance. It introduces cognitive debt, especially in relation to children and developing minds, as a concern created by dependence on commercial AI systems before adequate longitudinal evidence exists. It develops the idea of algorithmic epistemicide, where AI systems reproduce the historical exclusion of oral, ceremonial, embodied, relational, and non Western knowledge forms. It also examines the commercialisation of grief, where digital traces of the dead become material for simulation, intimacy, and profit. The expanded essay also adds a section on autonomous weapons and the accountability void created when lethal systems operate across military, corporate, technical, and legal chains of responsibility. This extension sharpens the paper’s broader claim that AI exposes gaps in responsibility wherever institutions rely on speed, delegation, proprietary systems, and ambiguous authority. The paper moves from inheritance, to audit, to authority crisis, to stress testing, and finally toward the structural questions raised by intelligence becoming infrastructure. Its case studies include political narratives, technology executives, AI companies, universities, research systems, banking, medical scanning, elite networks, welfare administration, insurance governance, grief technology, and military automation. Rather than asking whether AI should exist, the essay asks what legal, economic, institutional, epistemic, and moral architectures societies require now that machine cognition has entered public life, labour, governance, security, care, memory, and knowledge production.