Transformation of the Digital Tourism System: From Intelligent Agents to Agentic AI

The purpose of the paper is to examine the structural and functional transformation of the digital tourism ecosystem during the transition from traditional intelligent assistants, such as LLM-based chatbots, to a new paradigm of Agentic Artificial Intelligence. The subject of the present study is the qualitative transition from passive digital tools that rely on continuous human prompts to subjective systems characterised by autonomous planning, multi-agent coordination, and independent execution of complex service chains. The research methodology is based on a systems approach and organisational theory, using comparative analysis of artificial intelligence architectures to highlight the functional evolution of the industry. The methodological framework underpinning this study is predicated on the cyclical Action Research model, which encompasses planning, action, observation, and reflection. This model elucidates the iterative reasoning and self-repair capabilities of Agentic systems. Furthermore, the study employs the notion of multi-agent orchestration to model the interaction between specialised digital entities and proposes a conceptual model for assessing the socio-economic and ethical aspects of AI implementation, balancing human interests, environmental sustainability, economic profit and long-term development potential. The primary objective of this study is to provide a conceptual justification for the transition from auxiliary service tools to full-fledged digital management entities in tourism. The article provides a comprehensive overview of the conceptual features of Agentic AI, advances a functional model for its application in the tourism sector, and posits that these systems have the capacity to autonomously manage fragmented tourism services, including logistics, accommodation, and insurance, into a unified, personalised, and self-correcting service chain. The analysis determines that Agentic AI represents a fundamental change in the design of intelligent systems, going beyond the linear processing typical of traditional chatbots. The key conclusion is that Agentic AI serves as an integral element of the management subsystem, capable of implementing the full cycle of classic management functions, such as strategic planning, dynamic organisation through orchestration, and iterative control through feedback loops. The study concludes that Agentic Artificial Intelligence is defined not simply as a technological upgrade, but as a new paradigm of digital management integrated into the complex socio-economic processes of the global tourism industry, marking a transition from auxiliary service tools to autonomous digital workers.