Collaborative supervision strategies for risk issues of generative artificial intelligence in the tourism industry

This study investigates the collaborative supervision strategies of Generative Artificial Intelligence (GAI) in the hospitality and tourism industry, focusing on managing risks in GAI-based travel recommendations. Using evolutionary game theory and system dynamics, a tripartite model is developed, involving the government, GAI companies, and tourists. Five evolutionarily stable strategies (E1, E5, E6, E7, E8) are identified, showing how stakeholder behaviour evolves under different scenarios. The model operates through 'dual supervision, one implementation', with government regulation, tourist monitoring, and GAI self-regulation via AI or manual reviews. Simulations indicate that combined government and tourist involvement enhances self-regulation, with subsidies being more effective than penalties. While GAI companies may currently favour manual reviews, the potential for AI-driven reviews remains substantial. This framework provides practical insights for sustainable GAI governance in tourism.