Exploring the Application of AI/ML Systems in Fostering Regulatory Compliance and Audit in Oncology Trials

Abstract

This study investigates the integration of artificial intelligence (AI) and machine learning (ML) systems in regulatory compliance and audit processes within oncology clinical trials. Through a quantitative survey of 102 clinical research professionals, the research evaluates current adoption patterns, effectiveness, and challenges of AI/ML implementation in compliance monitoring. The findings reveal that while 47.1% of organizations currently use AI/ML tools for compliance, adoption remains fragmented and transitional. Natural Language Processing (19.6%) emerged as the most commonly used technology, primarily for documentation management. AI/ML systems demonstrated moderate effectiveness in detecting protocol deviations (28.4% rating them moderately effective) and reducing data discrepancies (46.1% reporting improvements). Risk assessment (28.4%) and compliance monitoring (28.4%) were identified as the most improved aspects following AI/ML adoption. However, significant barriers persist. Key challenges include data privacy concerns (23.5%), lack of regulatory clarity (21.6%), and technical complexity (20.6%). Only 41.2% of respondents believed their AI/ML tools were adequately validated for regulatory compliance, while 25.5% perceived these systems as lacking transparency.

The study revealed role-based differences in perceptions, with technical professionals expressing greater confidence in AI/ML systems compared to regulatory and quality assurance staff. Despite demonstrable benefits in error reduction (63.7% reported decreased human error) and audit efficiency, only 32.4% of respondents recommended wider adoption. This cautious stance reflects ongoing concerns about validation standards, explainability, and regulatory acceptance. The research concludes that while AI/ML offers significant potential for enhancing oncology trial compliance, successful integration requires clearer regulatory frameworks, improved system transparency, and cross-functional alignment between technical and compliance teams. These findings provide crucial insights for organizations navigating the complex landscape of AI/ML adoption in regulated clinical research environments.