Workers' Incentives and the Optimal Taxation of AI

TL;DR

This paper analyzes optimal tax policies in an economy with human labor, capital, and AI, finding that taxing AI becomes optimal when AI capabilities lead cognitive workers to switch to manual jobs.

Contribution

It extends existing dynamic taxation models to include AI, identifying the threshold at which taxing AI is optimal based on AI's substitution capabilities.

Findings

Optimal AI tax policy activates when AI substitutes for cognitive tasks.

Threshold for taxing AI depends on AI's capability to replace human cognitive labor.

Provides a framework for policymakers on AI-related taxation timing.

Abstract

We characterize the optimal tax policy in an economy with human manual and cognitive labor, physical capital, and artificial intelligence (AI). Extending the dynamic taxation setup of Slavik and Yazici (2014), we find that it is optimal to start taxing AI when cognitive workers start to consider switching to manual jobs. This threshold may be crossed once AI becomes sufficiently capable in substituting humans across cognitive tasks.