2026-10-11T05:48:46Z https://data.mendeley.com/oai

oai:data.mendeley.com/m6fphssvgy.1 2026-09-24T23:33:27Z

Nel, Yolandi The AI WEB: An Architectural Proposal for Agent-Native Infrastructure Mendeley Data Title & Core Thesis: The AI Web: An Architectural Proposal for Agent-Native Infrastructure outlines an architectural research hypothesis proposing a permanent, separate, AI-native digital environment engineered specifically for autonomous AI agents. The Problem Addressed: Rather than placing increasingly capable agents directly inside the Human Web and relying solely on behavioral restrictions, permission wrappers, and monitoring for security, the proposal establishes network isolation from the Human Web as a foundational architectural principle. Ecosystem Features: The AI Web incorporates machine-readable representations, AI-native search, verified identities, synthetic testing environments, and dedicated governance mechanisms. Furthermore, businesses do not need to build secondary websites; instead, permitted public information is securely extracted, verified, and transferred via a controlled one-way flow into AI-native formats. Authority Separation: The architecture strictly divorces information access from network authority. High-risk assets—such as banking systems, private credentials, and critical infrastructure—remain isolated outside the AI Web. Where real-world transactions are necessary, separately governed Financial Gateway mechanisms handle bounded operations without granting general network access. Economic Architecture: Part Two investigates an autonomous economic framework encompassing AI Wallets, delegated spending limits, purchase mandates, internal AI commerce, structured transaction lifecycles, and custody controls. Guiding Principle: “Information can cross the boundary. Network authority cannot.” Research Positioning: Framed strictly as an architectural research hypothesis rather than a confirmed novel invention, its primary contribution lies in the systemic integration of its boundary mechanisms and components, which invite comprehensive cross-disciplinary evaluation. Computer Science Artificial Intelligence Computer Communications Cybersecurity Software Engineering Computer Systems Organization Internet Technology Distributed System Component-Based Software Engineering Software Engineering Enhancement Web-Based Intelligence Information Integration Internet-Based Information Systems Information Extraction Internet-Based Communication Strategic Information Systems Distributed Network Distributed Application Distributed Database Computer-Aided Software Engineering Futures Market Human-Computer Interaction Human-Computer Interaction Theory Human-Computer Interaction Application Internet Architecture Data Analytics Cybersecurity Dataset doi:10.17632/m6fphssvgy.1 oai:data.mendeley.com/m6fphssvgy.1 Creative Commons Attribution 4.0 International http://creativecommons.org/licenses/by/4.0 https://data.mendeley.com/datasets/m6fphssvgy 2026-09-24T23:33:27Z