AI4Audit is AI4FinTech’s initiative to integrate cutting-edge artificial intelligence into the audit profession, transforming how auditors analyze data, assess risk, and ensure compliance. As audits grow in complexity and scale, AI4Audit explores how machine learning can enhance audit quality, efficiency, and transparency—bridging academic innovation with real-world audit practice.

This presentation is structured around two focused explorations of AI in auditing:
Graph Neural Networks for Anomaly Detection in Transaction Data
In the first part of the talk, Dr. Boersma will present how graph neural networks (GNNs) can be applied to model complex relationships in financial transaction data. By capturing the structure and flow of transactions as a network, GNNs enable auditors to detect subtle anomalies and patterns that traditional methods might miss. The talk also highlights the importance of explainability in these models to ensure auditability and trust.

Retrieve-Augment-Generate (RAG) for Disclosure Checking in ESG Audits
In the second part, George Sidiropoulos, Senior Researcher at KPMG, will present his work on using the RAG paradigm to automate the verification of corporate disclosures, particularly in the context of Environmental, Social, and Governance (ESG) reporting. His talk will cover how Document Question Answering (DocumentQA) models can retrieve and reason over large volumes of unstructured text to support ESG compliance, and how these models are evaluated and improved to meet evolving regulatory standards.