Ethical Challenges in Arabic–English Machine Translation: Toward a Human-Centered and Culturally Competent Approach

This article explores the serious ethical concerns surrounding machine translation (MT), particularly when applied to low-resource languages such as Arabic. While MT has immense influence on cross-lingual and cross-cultural communication, its ethical implications remain underexplored. The objective is to investigate key ethical issues in MT that are often overlooked by non-professional users and the public. The study reviews recent research to identify different types of biases in MT. It analyzes how these biases emerge in training data and algorithmic design and evaluates their cultural and linguistic impact. Practical suggestions are also offered for developing MT systems that are both ethical and culturally sensitive. Findings reveal multiple ethical issues associated with MT use for the Arabic–English language pair. These include gender bias, cultural insensitivity, and errors in translating context-dependent texts. In addition, MT use has negative consequences for translators, particularly in relation to job satisfaction and autonomy. Risks to confidentiality, especially in specialized texts, were also identified as a major ethical concern. The study emphasizes the importance of adopting a human-centered design approach. It calls for the active participation of Arabic-speaking communities, not as passive recipients of technology, but as co-creators of MT solutions. Addressing these concerns will foster the development of more reliable, ethical, and effective MT tools, advancing the field as a whole.