Generative artificial intelligence is now said to be reshaping the financial sector, emerging as a sophisticated tool in transforming how credit risk is being assessed. This, according to a report from OliverWyman which also noted that by moving further beyond traditional linear models, generative AI enables us to accurately analyze vast amounts of unstructured data — such as payment histories and digital behavior — in order to capture “subtle signals that enhance predictive accuracy.”
The technology not only aims to enhance the quality as well as depth of risk assessment but also improves upon more manual, time-consuming processes, particularly in handling complex corporate documentation. As a result of this, financial institutions are able to operate with more speed and efficiency in an increasingly competitive global environment.
The update from OliverWyman highlights how banking institutions have long depended on credit scoring models that work with linear relationships.
But generative AI’s ability to incorporate unstructured data “significantly increases the predictive power of these models.”
Tiago Rodrigues de Freitas, a partner and head of data and analytics Iberia at Oliver Wyman, and Ignasi Barri, global head of AI and data at GFT Technologies, emphasize that beyond simply improving quality and discrimination, generative AI “accelerates manual tasks, reduces human error and saves time.”
They also point out various challenges, such as the fast pace of technological change causing client hesitation, the “need for explainable AI models to build trust, and the importance of upskilling the workforce to adopt these solutions effectively.”
As an example of generative AI’s application, the professionals mention the Credit Risk Assistant developed by GFT and Oliver Wyman. This tool helps credit analysts efficiently process “diverse data sources across multiple languages.”
Looking ahead, they express confidence that AI will evolve from “merely supporting credit processes to becoming the central enabler of their re-engineering” — signaling a seemingly inevitable transformation in the financial services sector. But like with any new technology, careful human monitoring and involvement is still needed to ensure accuracy. These tech breakthroughs are most likely not meant to fully replace the need for human intervention. And will probably complement ongoing efforts to improve existing processes.
