Fairness of ChatGPT and the Role Of Explainable-Guided Prompts
This demonstrates LLMs' potential for data-efficient and fairer credit scoring, though it is incremental as it does not surpass existing ML methods.
The study examined whether ChatGPT could match traditional machine learning models in credit risk assessment, finding it achieved comparable performance using 40 times less data (20 vs. 800 data points) while improving fairness and reducing false positives.
Our research investigates the potential of Large-scale Language Models (LLMs), specifically OpenAI's GPT, in credit risk assessment-a binary classification task. Our findings suggest that LLMs, when directed by judiciously designed prompts and supplemented with domain-specific knowledge, can parallel the performance of traditional Machine Learning (ML) models. Intriguingly, they achieve this with significantly less data-40 times less, utilizing merely 20 data points compared to the ML's 800. LLMs particularly excel in minimizing false positives and enhancing fairness, both being vital aspects of risk analysis. While our results did not surpass those of classical ML models, they underscore the potential of LLMs in analogous tasks, laying a groundwork for future explorations into harnessing the capabilities of LLMs in diverse ML tasks.