Agentic Architecture of Information Systems — The AAM/5C Model: A Reference Model for Distributed Autonomy

Abstract

The technical leadership of large organizations faces what the book calls the cognitive complexity wall: at the scale of hundreds of interacting components, maintaining an Information System's dependency map and reacting at the pace of incidents increasingly exceed what a human team can deliver. This book argues that the answer is architectural: entrusting components with a share of perception, decision and coordination, without giving up control. It proposes AAM/5C (Autonomic Architecture Model / 5 Capabilities), a reference model that instantiates five Capabilities — Capture, Comprehend, Choose, Coordinate, Control — across seven logical layers. Its principle: intelligence proposes, but never acts without authorization. A single authorization rule separates the proposal of an act, the decision against a signed policy, the issuance of an act token and its verification by a deterministic policy enforcement point, placed on the execution path; a human organization writes, audits and revises the policies. The book owes most of its materials to autonomic computing, multi-agent systems, enterprise architecture and agent protocol security. Its contribution is a prescriptive reference model that articulates, under this single authorization rule, known mechanisms — separation of decision, issuance and enforcement, bounded authorization, the autonomic loop — and adds three constructs, all to be validated: the Validation step of the Distributed Autonomy Loop, the ephemeral authority surface and the cognitive dependency plane. As a conceptual volume, it reports no measurements. Two analytical statements are guaranteed by construction, under stated assumptions; the other benefits are design targets, framed as ten falsifiable hypotheses, with validation protocols and an economic model that can fail. The model targets large-scale Information Systems: depending on the scenario, its economic target H10 (net savings over total cost of ownership) is only reached from about 260 to 1,950 components. It does not target four families of systems, including critical embedded real-time systems.