Influence of racial bias in the use of facial recognition applied to access control: A critical analysis

Racial bias has been a persistent issue in facial recognition technologies, particularly within access control applications. This study aims to examine the widespread adoption of these technologies in the machine learning era, highlighting their integration into information security, cybersecurity, and data privacy frameworks. Despite their growing prevalence, the underlying datasets and algorithms frequently exhibit significant biases, disproportionately impacting individuals from marginalized racial groups. Through an extensive literature review, this research identifies critical gaps and proposes 14 targeted recommendations aimed at mitigating racial bias in facial recognition systems. These recommendations encompass diversifying training datasets, enhancing algorithmic transparency, and incorporating multidisciplinary teams to ensure ethical decision-making. The findings underscore the potential to improve both the equity and accuracy of these technologies, paving the way for more reliable and inclusive applications. By implementing the proposed measures, stakeholders can address ethical concerns, reduce discriminatory outcomes, and enhance public trust in the adoption of facial recognition for sensitive access control contexts. This critical analysis provides a roadmap for advancing fairness and accountability in artificial intelligence, fostering transformative impacts in the field.