We have provided feedback to the US Center for Audit Quality (CAQ) on its publication on generative AI (genAI) in financial reporting and audit. In our comment, we highlight genAI’s transformative potential across audit processes, talent and quality management while addressing new risks such as AI-driven biases and data integrity concerns.
Key points addressed:
- Use cases in audits: AI applications in audit and assurance services are on the rise. AI tools are increasingly used to extract insights from large datasets and detect anomalies. Audit firms are investing in genAI tools to improve the quality and efficiency of their services.
- Training needs: Enhanced, tailored training for audit and assurance professionals is necessary as technology advances. Professional bodies are reconsidering certification and continuing professional education requirements to ensure practitioners are equipped to handle the new technologies.
- Risk considerations: Auditors should understand regulatory frameworks and assess the risks brought by the use of AI-based systems in corporate reporting and audit. Key concerns include data reliability, algorithmic biases and confidentiality issues, and require particular attention.
- Regulatory developments in EU: The EU AI Act, effective by 2026, introduces new rules for AI systems development and deployment in the EU. It follows a risk-based approach and mandates that companies document AI governance. Additionally, the European Single Access Point (ESAP) will provide machine-readable financial and sustainability information data, which can be used to train AI models in reporting, auditing and training programmes.
- Looking to the future: Supervisory approaches and professional standards must support technological innovation. Players in the AI ecosystem should consider the specific needs of small and medium-sized practices (SMPs) to maintain a resilient audit market. Companies and audit firms should assess environmental and social impacts of AI, including energy consumption and algorithmic biases to promote responsible AI development.
Find the full feedback in the download section.
