Work

Fair and square? Evaluating fairness of LLM-generated synthetic datasets

Year: 2025

Type: article

Abstract: Context. Machine Learning (ML) is driving advancements across various industries, including healthcare, finance, and entertainment, but it also raises significant ethical concerns, particularly regarding fairness. Biases in training data can lead to unfair outcomes, perpetuating or even amplifying existing biases. While the software engineering community has developed various bias mitigation techniques, our research identifies two key limitations: (1) these methods often address bias only in later stages of development, such as after data collection or model training; and (2) they may only mitigate bias and possibly not fully address it, as the root issue often lies in the data itself. Objective. In this paper, we explore an alternative approach to mitigate unfairness: synthetic data generation, which involves creating artificial datasets that mimic the statistical properties of real-world data. We aim to assess how this approach can contribute to generating data that positively impacts the trade-off between performance and fairness by creating datasets that reduce the influence of real-world biases through synthetic feature generation. Methods. To this end, we conducted an empirical study comparing ML models trained on synthetic datasets generated by large language models to ML models trained on real-world data, evaluating performance and fairness indicators. Results. Our results demonstrate that models trained with synthetic data, particularly those generated using simpler prompts, can achieve competitive performance while enhancing fairness. Conclusion. Our work suggests that synthetic data generation may be a viable approach to addressing fairness requirements in ML systems.

Source: Information and Software Technology

DOI: 10.1016/j.infsof.2025.107980