Generative AI, Work and Risks in Cultural and Creative Industries

Abstract

This paper analyses the widespread transformation of work in the cultural and creative industries associated with exposure to generative AI. We provide an empirical methodology through a task-based framework to measure direct and indirect exposure to generative AI and its associated risks. We show that the use generative AI at work requires new AI risk management abilities. These risks relate to privacy, cybersecurity, breach of professional standards, bias, misinformation, accountability, and intellectual property. Our approach contributes to developing a new perspective on the inseparability between generative AI and risk exposure as a driver of job transformation. Ignoring this aspect in previous AI-exposure measures has distorted the understanding of the generative AI adoption and job transformation. We map the exposure to generative AI and these eight risks across 593 specialist skills of 126 occupations in the cultural and creative industries, relying on the Australian Skill Classification of December 2023 and innovative synthetic data methodology. We identify five job transformation zones that highlight the intensity, timing, and nature of risks to guide the implementation of risk mitigation and upskilling strategies.

Keywords: Generative AI, AI risk, Synthetic Data, Technology exposure, ChatGPT, LLM, Work, Jobs, Technological change, Cultural and Creative Industries

JEL Classification: J24, O33, Z19