Proactive Hybrid Recommendation Advertising in Generative AI Text Dialogues: An AI-Initiated Text Advertising Format Integrating Organic and Sponsored Recommendations

The emergence of generative artificial intelligence (AI) is transforming digital communication by enabling users to obtain information, assistance, and recommendations through natural-language conversations. As generative AI platforms explore advertising-supported business models, an important research opportunity concerns the development of advertising formats that are native to text-based AI interactions rather than simply adapting conventional advertisements to conversational interfaces. This paper proposes Proactive Hybrid Text Advertising (PHTA), a conceptual advertising framework specifically designed for text-based generative AI dialogues. Under PHTA, a generative AI system proactively initiates and presents a unified textual recommendation list without waiting for the user to explicitly request product recommendations, service suggestions, or advertising content. The list integrates two distinct categories of recommendations: organic text recommendations, independently selected according to their relevance and potential usefulness to the user, and sponsored text recommendations, included through identifiable paid advertising arrangements. The defining contribution is the conceptualization of the entire AI-initiated hybrid recommendation list as an independent text advertising unit, rather than treating advertising exclusively as an isolated sponsored sentence, a paid recommendation appended to an ordinary AI response, or a sponsored entry within a user-requested recommendation list. The proposed framework introduces conversational-context analysis, proactive advertising-opportunity identification, organic recommendation generation, sponsored recommendation selection, hybrid-list construction, text-based sponsorship disclosure, and user-control mechanisms. It further establishes a separation between organic recommendation selection and advertiser-funded placement to reduce the risk that commercial incentives silently determine the AI system's independent recommendations. PHTA is designed to support contextual relevance, transparent commercial classification, conversational continuity, privacy protection, and user autonomy. This paper develops the conceptual foundations, architectural components, operational workflow, design requirements, evaluation propositions, and ethical considerations of the proposed format. It distinguishes the framework from existing research on conversational advertising, sponsored LLM responses, and proactive conversational AI by focusing on the combined mechanism of AI-initiated presentation and a unified organic–sponsored text recommendation list. No experimental results, validated performance improvements, or demonstrated commercial advantages are claimed. To the best of the author's knowledge, the specific combination and formalization of these mechanisms as a distinct proactive hybrid text advertising format for generative AI dialogues constitutes a research proposal warranting further systematic investigation.