The Effect of Artificial Intelligence Intensity on Audit Quality: Evidence from Accounting Interns in Indonesia

Abstract

This study examines the effect of artificial intelligence (AI) intensity on audit quality among accounting interns at registered public accounting firms (KAP) in Indonesia. The accelerating deployment of AI technologies — including machine learning, robotic process automation (RPA), data analytics platforms, and anomaly-detection algorithms — has fundamentally restructured audit practice; yet the individual-level implications for early-career auditors operating within AI-augmented environments remain empirically underexplored. Anchored in the Technology Acceptance Model (TAM) and Agency Theory, this study adopts a quantitative, cross-sectional design, collecting primary data from 31 purposively sampled accounting interns through a validated five-point Likert-scale questionnaire. Data were analysed using simple ordinary least squares (OLS) regression in SPSS. Descriptive statistics reveal a moderate level of AI intensity (M = 3.23, SD = 0.71) and a moderately high level of perceived audit quality (M = 3.65, SD = 0.69) within the sample. The OLS regression model is statistically significant (F = 9.636, p = 0.004, R² = 0.249), and the AI intensity coefficient is positive and significant (B = 0.485, β = 0.499, t = 3.104, p = 0.004), indicating that each unit increase in AI intensity is associated with a 0.485-unit improvement in perceived audit quality. These results confirm H1 and provide micro-level quantitative evidence that higher AI integration enhances audit outcomes among interns. Concurrently, the study highlights the latent risk of overreliance: uncritical acceptance of AI-generated outputs may erode professional scepticism — a competency that remains irreplaceable in high-stakes financial reporting verification.