The rapid evolution of Artificial Intelligence (AI) brings diverse applications in optimization, recommendation, and prediction. While AI ideally reduces human biases, it often automates inequality, necessitating measures to address gender biases. The Canadian government’s Algorithmic Impact Assessments (AIA) is a commendable step, mandated to evaluate AI’s societal repercussions. However, this paper, adopting a feminist perspective, argues for refining AIA. To effectively combat gender bias, a holistic approach is vital, including diverse perspectives in AI development, regular audits, enhanced AIA inclusivity, and comprehensive diversity training for developers. This multifaceted strategy aims not only to identify and rectify biases but also to foster equitable and fair AI systems, preventing the perpetuation of inequality in technology.