Incentivizing the Content Creator When Tasks Can Be Delegated to Artificial Intelligence

Abstract

The advent of Artificial Intelligence (AI) tools provides substantial resources for human creators when contributing content to platforms. However, these tools also introduce challenges, as human creators may excessively delegate tasks to AI to reduce costs. This paper focuses on the scenario in which AI does not fully replace human creators for every task and introduces a game-theoretic model to explore how platforms should adjust the revenue-sharing level to incentivize human creators. Our model reveals that, for a representative human creator, if delegation effectively enhances overall output, the platform should reduce the revenue-sharing level, as AI can partially substitute the human creator. Conversely, if delegation does not sufficiently improve overall output, the platform should increase the revenue-sharing level to encourage human effort on tasks that AI struggles with. Furthermore, we demonstrate that the introduction of strong AI tools could be detrimental to the platform, as it may lead the human creator to excessively shirk effort—a problem that cannot be resolved by merely increasing the revenue-sharing level. Interestingly, if the platform can adjust AI strength endogenously, the optimal revenue-sharing level remains unchanged from the no-AI scenario. This finding significantly simplifies the platform’s revenue-sharing decision-making process in the AI era. Moreover, we show that excessively strong AI can harm other stakeholders as well and diminish overall social welfare. Finally, we extend our analysis by showing that allowing AI to augment human creators at the task level does not qualitatively alter our key findings, and that imposing charges on AI usage fails to prevent excessive automation due to the platform’s self-interested incentives.