Langues & Cultures
Volume 6, Numéro 4, Pages 09-27
2025-12-25

Ethical Risks And Safeguards In Generative Ai: Managing Impacts On Cognition And Decision-making.

Auteurs : Boumediene Houda . Ouarniki Ouafa .

Résumé

Abstract: This study concentrates on the double-edged nature of AI as an innovation enabler and a source of ethical issues such as hallucinations, disinformation, and output variability and brittleness of large language models (LLMs) to prompts and context. As AI increasingly informs human decision-making and can amplify existing cognitive biases, it also provides the opportunity for improvement. The paper calls for a firm moral foundation for the development of AI to guide innovation toward societal values. Proper use of AI requires developers, researchers, policymakers, and end-users to work together. Ongoing end-user engagement and training are essential in managing risks as well as establishing the best practices. Legal frameworks will be needed to keep up with the faster growth of AI. Openness, ethics, and transparency must be given priority to usher in human development and ensure that AI drives innovation according to ethical principles.

Mots clés

Artificial intelligence; ethical framework; large language models (LLMs); sharing of experiences ; societal values .