Position: Can Agentic AI Make Gig Economy More Fair ?

Cheng Chen, Chin Wang Cheong, Jing Li, William K. Cheung, Ivor Tsang

Abstract: The advent of Agentic Artificial Intelligence System (AAIS) is poised to revolutionize the gig economy by driving growth, expanding the workforce, and automating complex processes. AAIS achieves its performance by breaking down complex tasks into subtasks, leveraging multi-process frameworks to tackle intricate systems effectively. This approach often requires the use of multi-modal or multi-language models, which are inherently susceptible to algorithmic biases. Furthermore, managing unstructured data—spanning natural language processing, images, videos, and meta-datasets—is indispensable for building robust AAIS. However, every aspect of unstructured data engineering is riddled with biases, which further amplifies the potential for unfair outcomes in the gig economy. Therefore, \textbf{We argue that deploying AAIS without addressing these systemic biases will inevitably compromise fairness in the gig economy.} To mitigate these challenges, we advocate for the urgent introduction of fairness assessment and mediation mechanisms tailored to AAIS, which are critical for fostering fairness in gig economy.

Primary Area: Social, Ethical, and Environmental Impacts

Keywords: Fairness in agentic AI system, Fairness in Gig Economy

Submission Number: 45