The Demand Factor in the Age of AI: fiscal principle, measurement architecture, and financing horizon

If artificial intelligence advances toward structural substitution of human labor — one of two trajectories that dominate the current debate — it could erode the wage channel that for more than a century distributed purchasing power and sustained effective demand. Economies would then generate unprecedented productive capacity while employment-based wage income declined in relative terms, opening a structural gap between what can be produced and what the market can realize. This work argues that, in such a scenario, the human being does not disappear from the economic circuit: the function changes. From a central producer the human being would become the indispensable bearer of the demand factor — the human capacity to absorb production through purchasing power, which would transform automated production into realized market value. From this Copernican turn, the IRIS framework articulates three interconnected instruments: the Market Access Tax (MAT), which captures the uncompensated incidence of AI on domestic production; the Market Access Right (MAR), which extends the same logic to access from outside the jurisdiction; and the Market Access Dividend (MAD), its distributive counterpart. The architecture is reinforced by the institutional efficiency dividend generated by the new efficiency of the AI-assisted public sector. Entitlement to the MAD is recognized in the social body that sustains the effective demand of the market — not as an assistance category defined by need, but as factor compensation for a collective contribution. By principle, that title extends to the citizenry as a whole, whose economic function as consumers should not, by definition, place its structural position below the subsistence threshold. The paper further develops a transitional measurement architecture that evolves from technical, economic, and declarative proxies toward an enterprise interpretive agent assisted by AI. And it confronts squarely the structural limit of the fiscal horizon: even a complete MAT-MAR-MAD system, enhanced by institutional efficiency, may prove insufficient to rebuild broad effective demand in a deeply automated economy. Linking automation, longevity, pension sustainability, and the care economy, IRIS is configured as a bridge architecture: it buys institutional time for an orderly transition toward new forms of human work while preventing the collapse of demand. The question it leaves open is no longer merely fiscal: it concerns how to articulate institutionally a productive capacity expanded by AI with an effectively distributed purchasing power.