Generative AI in enterprise IT implications for audit evidence, assurance, and governance

The increased adoption of generative artificial intelligence (AI), in particular large language models, across enterprise IT environments is starting to disrupt the process by which information is produced, recorded and consumed for decision making. While much of the existing discussion emphasizes performance and efficiency gains but less attention has been given to the implications for audit evidence, assurance and governance. This paper examines how AI-assisted content generation affects fundamental audit concepts such as evidence reliability, data provenance and audit trail integrity. Drawing on common enterprise use cases, the study explores areas where traditional audit assumptions may no longer hold, especially when AI-generated outputs are blended with human judgment. The analysis highlights practical challenges auditors face in tracing the origin of information, assessing the sufficiency of evidence and determining accountability for AI-influenced decisions. It also considers how existing governance and control frameworks appear to lag behind actual organizational practices. Rather than proposing entirely new audit models, the paper discusses pragmatic adaptations that IT auditors and governance professionals can apply within current structures. These include revised evidence classification approaches, enhanced logging and documentation expectations, and clearer governance policies for acceptable AI use. The paper aims to support auditors in maintaining assurance in environments where audit evidence is increasingly shaped by generative systems rather than exclusively by human actors.