Unveiling the Future of Translation: A Qualitative Study on the Benefits, Challenges, and Future Venues of Translation with AI
Volume 16, Issue 1, November 2025, Pages 251-282
https://doi.org/10.22059/jolr.2025.403095.666941
Asghar Moulavinafchi, Reyhane Sadat Shadpour, Yasir Hazim Qasim Qasim
Abstract Against the backdrop of rapid advances in artificial intelligence (AI), this qualitative study investigates how professional translators perceive the benefits, challenges, and likely future trajectories of AI-assisted translation. We conducted semi-structured Zoom interviews with 20 translators (≥3 years’ experience) from diverse specializations and professional contexts. Interviews were transcribed verbatim and analyzed thematically following Braun and Clarke’s procedures, with code-checking by an experienced colleague to enhance credibility. Participants reported four prominent benefits of AI-enabled workflows: (1) efficiency gains through pre-drafting and automation of repetitive segments; (2) cost-effectiveness for large or time-sensitive projects; (3) improved consistency and terminological control, particularly in technical domains; and (4) smoother collaboration via shared memories and glossaries. Equally salient, however, were perceived risks and frictions: AI’s persistent difficulty with cultural-pragmatic nuance and rhetorical effect; over-reliance that can deskill practitioners and inflate client expectations; post-editing fatigue stemming from unnatural or ambiguous output; ethical concerns regarding data privacy, confidentiality, authorship, and job security; and technical limitations in low-resource languages or complex genres. Synthesizing convergent and dissonant evidence, we propose a human-in-the-loop “personalization and productivity” model in which AI serves as a co-pilot embedded within a human-led pedagogical and professional envelope. Within this model, human expertise remains decisive for problem framing, audience adaptation, quality assurance, and value-laden decisions. Implications include (a) targeted professional development in AI literacy, post-editing strategies, and bias awareness; (b) curricular redesign in translator education to foreground hybrid competencies and critical evaluation of AI outputs; and (c) organizational and sector-level policies for data governance, transparency, and fair remuneration. Limitations concern the modest, predominantly Iranian sample and reliance on self-report interviews; future research should incorporate longitudinal designs, task-based performance data, and comparative studies across languages and domains. Overall, our findings caution against both technological pessimism and automation hype, arguing instead for calibrated human–AI collaboration that leverages automation where it adds value while safeguarding cultural depth, ethical integrity, and the creative agency that defines professional translation.

