Proportionate Quality Management Systems for Agile AI Microenterprises : A replicable synthesis process and implementation roadmap to prepare for the EU AI Act

Sammanfattning

Regulation (EU) 2024/1689 (the AI Act) requires providers of high-risk AI systems to operate a documented quality management system (QMS) under Article 17. Microenterprises face this obligation in full whilst lacking the governance capacity of larger firms. Article 63 grants a simplified fulfilment pathway but does not define how it is to be operationalised, and the agile culture typical of early-stage AI providers stands in tension with the formalised processes the standards presuppose. No published method synthesises Total Quality Management (TQM), ISO 9001:2015, ISO 42001:2023 and Article 17 into a proportionate management system for such organisations. This study develops and documents a replicable process for synthesising a proportionate QMS for an AI microenterprise and shows how it can be sustained alongside existing agile practices. The theoretical framework combines the TQM cornerstone model and the PDCA cycle, the two ISO standards and Scrum. The study is a qualitative, normative single-case study employing a five-stage procedure: normative document analysis, an empirical baseline from interviews and observations, hybrid coding with a four-state stage classification, framework synthesis through three selection criteria, and an implementation roadmap. A fifty-element inventory yielded forty-eight included elements (twenty full, three simplified, twenty-five beyond minimum) and two deferred, sustained through six integration points within Scrum events and adjacent reviews. Proportionate implementability favoured exceeding the Article 17 minimum at microenterprise scale, and Article 63 simplification operates on administrative burden, not substantive scope. The framework remains defensible irrespective of regulatory amendment.