普遍的共識指出嚴重的所得稅侵蝕風險。核心經濟模型表明人工智慧的採用與政府收入之間存在負相關：隨著人工智慧取代高薪職業，即使是邊際百分比的失業率增加，也將產生不成比例的龐大財政赤字。這種統計上的分歧——其特徵為稅收下降與社會福利支出攀升同時發生——可能會引發被稱為「人工智慧稅收缺口」的多年結構性短缺，進而潛在地威脅數十年的主權穩定。
為了減輕這種預期的收入崩潰，經濟學家提出了4種主要的稅收策略，針對那些從人工智慧整合中獲益最多的人。這些框架包括實施財富或奢侈稅、增加基準消費稅、提高企業稅率，或針對人工智慧運算消耗與代幣建立新型徵費。然而，正如 David Ramli 所分析的，這些財政政策的實施面臨著一個普遍的行為經濟學挑戰：大眾對任何總體稅收增加的內在抗拒。
Last December, 50 finance and technology experts from organizations including Google DeepMind, the Rand Corporation, and the Federal Reserve System convened at the International Monetary Fund to model the economic impacts of artificial intelligence. Divided into 6 distinct working groups under the Chatham House Rule, their primary objective was to analyze potential statistical shifts in global gross domestic product, government spending, and taxation structures.
The prevailing consensus indicated a significant risk of severe income tax erosion. The core economic model suggests an inverse correlation between AI adoption and government revenue: as artificial intelligence displaces high-paying occupations, even a marginal percentage increase in unemployment will generate a disproportionately large fiscal deficit. This statistical divergence—characterized by simultaneously decreasing tax yields and escalating social welfare expenditures—could precipitate a multiyear structural shortfall termed the AI tax gap, potentially threatening sovereign stability over several decades.
To mitigate this projected revenue collapse, economists have proposed 4 primary taxation strategies targeting those benefiting most from AI integration. These frameworks include implementing wealth or luxury taxes, increasing baseline consumption taxes, elevating corporate tax rates, or establishing a novel levy on AI computing consumption and tokens. However, the implementation of these fiscal policies faces a universal behavioral economic challenge, as analyzed by David Ramli: the inherent public resistance to any aggregate tax increases.