The development of autonomous AI agents and LLM-based multi-agent systems creates conditions for the emergence of organisational structures that are not entirely predefined, but are formed through interaction between the agents. This paper proposes the Self-Organising AI Community Lifecycle Model, which describes the development of a self-organising AI community through six organisational states – from an unstructured population to adaptive self-organisation. The model distinguishes the functional architecture from the emergent organisational stages and defines measurable conditions for transition between them. A framework is also proposed for subsequent experimental evaluation of the model under varying degrees of predefined organisation.

autonomous AI agents,multi-agent systems,self-organisation,AI communities,emergent behaviour,lifecycle model