Inteligencia artificial, discurso de odio y libertad de expresión: estándares comparados de la Corte IDH y el TEDH

Abstract

Introduction. This paper analyzes the legal and foundational framework of algorithmic moderation based on the antinomic tension between freedom of expression and hate speech. The relevance of this analysis lies in the need to unravel, from a dogmatic perspective, the categorical disruption caused by generative artificial intelligence as a mere technical tool within international law. Methods. A critical and comparative review of recent case law from the Inter-American Court of Human Rights and the European Court of Human Rights was conducted. Results. The findings reveal a divergence in standards: while the European system consolidates a due diligence model that extends criminal liability to account holders and platforms, the Inter-American system maintains a structural resistance to any form of preventive censorship in debates of public interest. Conclusion. Based on the arguments put forth by Cotino Hueso and Llansó, it can be concluded that responsibility for this regulatory codification in the digital ecosystem lies directly with social media platform owners and with individual users who violate the rules, rather than with the capabilities of AI itself. Delegating judgment to a binary algorithm creates opacity and a form of “hygienic cleansing” that sacrifices democratic pluralism by subordinating fundamental rights to a logic of neutralizing linguistic harm, devoid of legal hermeneutics.