Biases within Artificial Intelligence Algorithmic Models in the Insurance Sector: The Need for Ethical Principles to Ensure the Protection of Policyholders’ Human Rights

Abstract

This article analyzes whether the adoption of ethical principles contained in CONPES 4144 of 2025 can be an effective way to guarantee equity, transparency and accountability in the use of artificial intelligence (AI) in the Colombian insurance sector in the face of the biases that algorithmic models may cause. The research is developed from an analytical approach, under a legal-normative and ethical perspective, through a documentary review of public policies, international standards and academic literature. The conceptual framework is based on the principles of algorithmic governance, non-discrimination and explainability, applied to highly regulated contexts.