From NVIDIA:

The digital frontier has become a hotbed for financial fraud, with credit card losses expected to reach $43 billion by 2026. Generative AI is being used to perpetrate financial fraud in various ways, from credit card theft to phishing for personal information. Financial services firms are turning to AI for fraud detection to combat these crimes in real time. The use of generative AI in financial fraud is a growing concern, as sophisticated technology can be exploited for malicious purposes. Fraudulent activities such as voice authentication and phishing scams are being committed with the assistance of generative AI. Financial institutions are leveraging generative AI to combat transaction fraud and improve accuracy, ultimately reducing financial losses and compliance risk. NVIDIA offers tools and platforms to help enterprises deploy generative AI for fraud detection, ensuring that AI-powered applications are secure and accurate. The financial services sector is also using AI-driven applications to enhance identity verification for regulatory compliance and cost reduction.

Graph neural networks (GNNs) are being embraced for their ability to reveal suspicious activity, helping financial institutions identify previously unknown patterns of fraudulent behavior. NVIDIA has developed an alliance with the Deep Graph Library team and the PyTorch Geometric team to provide a GNN framework containerized offering that includes the latest updates and NVIDIA RAPIDS libraries. GNNs enable financial institutions to track complex and longer transaction chains used by financial fraud perpetrators in attempts to obscure their tracks. GNNs are crucial for detecting financial fraud patterns at a massive scale and can be trained on unsupervised or self-supervised tasks. These techniques help developers pretrain models without labels and produce strong graph representations, ultimately aiding in the detection of financial fraud.



Read more: How Is AI Used in Fraud Detection?