Interpretable AI: Enhancing Transparency and Fairness in Decision-Making
Year: 2025
Type: article
Abstract: Interpretable Artificial Intelligence aims to make machine learning models more transparent, interpretable, and accountable, addressing the ”black box” nature of traditional AI systems. As AI plays a critical role in high-stakes domains like healthcare, finance, and autonomous systems, ensuring trust and fairness in decision-making has become essential and this paper also explores key techniques in AI. This study adopts a mixed-methods approach to analyse, evaluate, and compare XAI techniques across key domains. This research examines a three-phase approach in XAI, focusing on exploring different methods, evaluating their impact in real- world applications, and analysing the trade-offs between interpretability and model performance. XAI enhances transparency, bias detection, and user trust, but still it faces challenges, such as the trade-off between interpretability and accuracy, as well as computational complexity. Future studies focus on improving model interpretability, enhancing human-AI interaction, and promoting fairness in AI-driven decisions.ncy.
Authors: Rajeev Nair, Rosemol Thomas, Sandra MV, Ms. Siji K B
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4408527464