Work

AI-to-AI Communication: Toward a New Paradigm of Intelligence Search

Year: 2026
Type: preprint
Abstract: The rapid development of artificial intelligence has transformed the way people search for and interact with information. Traditional search engines primarily retrieve documents, web pages, and other information sources, while modern large language models can interpret questions, reason over information, and generate answers. This paper explores a possible next step: a search paradigm in which AI agents communicate with one another to search for, compare, evaluate, and synthesize answers. This paper proposes an AI-to-AI communication paradigm for intelligence search, extending information retrieval beyond the conventional interaction between a user and a single AI model to a collaborative process involving multiple specialized AI systems. Rather than merely collecting or aggregating their responses, the framework enables participating systems to independently investigate a query, retrieve relevant evidence, exchange findings, and critically assess one another's contributions. The resulting intelligence is evaluated and integrated to construct a well-supported response. A key distinction of the proposed architecture is its shift from answer-level agreement to claim-level verification. Instead of treating each AI-generated response as a single unit, the architecture decomposes responses into individual claims, independently retrieves supporting evidence, evaluates source independence, and accounts for dependencies among potential errors. This process enables the final synthesis to consider not only the agreement among AI systems but also the evidential support, credibility, and potential dependence underlying their claims. The numerical results provide initial evidence that the proposed framework can support a new architectural layer between AI systems, dedicated to discovering, coordinating, evaluating, and verifying distributed machine intelligence. By enabling structured interaction and evidence-based assessment across multiple AI systems, this layer offers a foundation for moving beyond isolated AI capabilities toward more coordinated and reliable forms of machine intelligence. This paper proposes an AI-to-AI communication paradigm for intelligence search, extending information retrieval beyond the conventional interaction between a user and a single AI model to a collaborative process involving multiple specialized AI systems.
Language: English
Open Access status: green
OpenAlex ID: W7222843668
Cites: 10
Cited by:
Related to: 10