Gender Disparities in Public Salary and Recruitment Datasets: A Descriptive Analysis with Implications for AI-Based Bias Detection

Workplace bias and gender disparities are barriers to professional environments. Artificial Intelligence (AI) has the potential to detect and mitigate these inequalities, yet raises a risk of amplifying them when models are trained on skewed data. This paper presents a statistical analysis of existing gender biases using recruitment and salary datasets available in the public domain. The study investigates differences in hiring outcomes and compensation by gender and education levels using Orange Miner for data visualization and exploratory data analysis (EDA). The paper conceptualizes these findings with women's empowerment schemes of the Indian government, highlights inequalities present in historical data, and examines the role of AI models in detecting bias and disparities. It emphasizes the requirements of designing ethical AI systems to ensure equality, transparency, and a supportive environment for women.