Audit Trails and AI Transparency: Regulatory Compliance under the EU AI Act

Audit trails serve as pivotal technical mechanisms for ensuring transparency and traceability, both of which are critical for compliance with the European Union's AI Act. This paper provides an analysis of the regulatory imperatives regarding operational transparency, automated event logging, and human oversight within the context of high-risk artificial intelligence systems. It scrutinizes the fundamental technical attributes of audit trails, the challenges inherent in their implementation, and the organizational and institutional prerequisites necessary for their effective integration. Furthermore, the study presents indicative tools and technological solutions—such as the ELK stack, Apache Kafka, and explainability techniques like SHAP and LIME—demonstrating their capacity to support compliance, transparency, and oversight. The analysis suggests that these mechanisms do not merely facilitate regulatory adherence but also enhance operational reliability, accountability, and the continuous monitoring of artificial intelligence systems throughout their lifecycle.