Ethical Implications of Artificial Intelligence in Fact-Checking: Insights from Fact-checkers and Scholars in Spain

Abstract

This study examines how fact-checking professionals, experts, and scholars perceive the use of artificial intelligence (AI) in information verification processes, focusing on its utility, ethical implications, and the existence of codes governing its application. A qualitative approach was adopted through 16 semi-structured interviews with Spanish fact-checkers and academics conducted between January and June 2025. The analysis carried out using Atlas.ti, employed inductive coding to identify patterns and perceptions regarding AI in verification. Findings indicate that AI enhances efficiency in the pre-verification stage and in the subsequent dissemination of verified content. However, during the core verification phase, its autonomy remains limited, and automated systems exhibit low reliability. The lack of neutrality in commercial tools used for verification generates skepticism among interviewees, who call for greater transparency in technological development. Additional risks include algorithmic bias, insufficient datasets, challenges in ensuring source reliability, and opacity in processes, all of which heighten ethical concerns about the accuracy of results. The study concludes that AI should be understood as a complementary tool within a human-in-the-loop model, rather than a substitute for professional judgment. Furthermore, in response to demands for transparency, the research underscores the need to promote the principle of explainability and to develop specific ethical codes ensuring traceability, neutrality, and protection against bias, aligned with international standards and societal expectations for trustworthy information