Answers are grounded in this paper and cite their sources.
Generative artificial intelligence has progressively evolved from a primarily reactive conversational technology into an interactive participant capable of supporting collaborative, multi-party, and context-sensitive communication. However, the ability to participate in a conversation and the ability to initiate the inclusion of additional participants represent two distinct interaction capabilities. Existing research has examined multi-party conversational engagement, conversational agent participation, proactive AI intervention, and human-controlled group interaction. Nevertheless, these research directions do not, individually, establish a unified invitation-oriented architecture in which both the generative AI system and the human user can independently initiate invitations, operate under different invitation permissions, and coordinate the inclusion of additional participants throughout an ongoing conversation. This paper introduces Generative AI Conversation Invitation (GACI), a unified conceptual framework that formalizes participant invitation as an independent and configurable capability of generative AI communication systems. The framework encompasses bidirectional invitation, AI-initiated invitation, user-initiated invitation, unidirectional invitation, context-aware invitee recommendation, invitation timing, purpose specification, permission management, invitation delivery, acceptance and rejection, withdrawal, and post-admission conversational continuity. It further distinguishes the authority to initiate an invitation from the authority to approve participation, ensuring that proactive AI capabilities do not automatically override human consent or communication boundaries. The paper establishes a conceptual foundation, interaction model, operational lifecycle, system architecture, design principles, and evaluation agenda for invitation-aware generative AI systems. To the best of our knowledge, this paper presents the first unified conceptualization specifically centered on bidirectional and unidirectional invitation authority between generative AI and human users as a distinct capability within ongoing human–AI communication. This contribution is intended to establish conversation invitation as an independent research and system-design dimension rather than treating it merely as a secondary feature of conventional group-chat functionality.
No takes yet. Share an insight, caveat, or question.
Jincheng Zhang (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: