Generative AI Conversation Request: A Unified Consent-Gated Framework for Bidirectional and Unidirectional Communication Initiation Between Humans and Artificial Intelligence
Year: 2026
Type: preprint
Abstract: The development of generative artificial intelligence (AI) has substantially expanded the possibilities of communication between humans and intelligent computational systems. Contemporary conversational interfaces, however, predominantly follow a response-oriented interaction model in which a user initiates a conversation and an AI system responds, or a proactive AI system initiates communication through notifications, suggestions, and unsolicited conversational interventions. These interaction patterns do not necessarily provide a unified mechanism through which one communication participant must formally request permission from another participant before establishing a new conversation. This paper introduces Generative AI Conversation Request (GACR), a unified conceptual framework for request-based conversational initiation in human–AI and AI–AI communication. The framework establishes a fundamental distinction between requesting a conversation and establishing a conversation: submitting a request does not itself authorize the commencement of a formal conversational interaction. Instead, the intended recipient must first accept the request, after which the conversation can be established under the permissions associated with that acceptance. GACR supports both bidirectional and unidirectional communication initiation, including human-to-AI requests, AI-to-human requests, and AI-to-AI requests. It further introduces a Consent-Gated Conversational Establishment (CGCE) mechanism that formalizes request submission, delivery, pending status, recipient evaluation, acceptance, rejection, expiration, withdrawal, and conversation establishment. The proposed framework also considers request metadata, communication scope, temporal constraints, permission granularity, identity verification, and revocation. Unlike approaches that primarily investigate proactive conversational behavior, conversational turn-taking, or agent communication after an interaction protocol has been established, GACR places the request for permission to establish a new conversation at the center of the interaction lifecycle. To the best of our knowledge, this paper presents the first unified conceptual formulation specifically centered on consent-gated conversation requests across human–AI and AI–AI participants, while recognizing that related permission mechanisms, proactive dialogue models, and multi-agent communication protocols have been studied previously. This work establishes a conceptual and architectural foundation for future implementation, interoperability, and empirical evaluation of request-based generative AI communication.
Author: Zhang, Jincheng
Language: English
Topic: Speech and dialogue systems
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7235573036
