Volume 15 Issue 06, June 2025
Impact factor: 2019: 4.679 2020: 5.015 2021: 5.436, 2022: 5.242, 2023:
6.995, 2024 7.75
http://www.internationaljournal.co.in/index.php/jasass
102
THE ROLE OF ARTIFICIAL INTELLIGENCE IN TRANSLATION
Zuparova Lobar Karimovna
Senior Lecturer
Uzbekistan State World Languages University
Abstract:
This article explores the role of artificial intelligence (AI) in translation through a
qualitative, sociotechnical perspective. It examines how AI tools impact translation efficiency,
accuracy, cultural sensitivity, and professional roles. Findings reveal a complex interplay where
AI enhances speed but struggles with nuanced meaning and cultural context, requiring ongoing
human post-editing and oversight. The study highlights ethical concerns around bias and trust,
and emphasizes that AI is reshaping rather than replacing the translator’s role. Implications for
practice, policy, and future research are discussed.
Keywords:
Artificial intelligence, translation, sociotechnical systems, post-editing, cultural
sensitivity, trust, professional roles, qualitative research
Introduction
In recent years, artificial intelligence (AI) has increasingly influenced various aspects of human
communication, with language translation being one of the most prominent domains. From real-
time speech interpretation apps to sophisticated machine translation systems like Google
Translate, DeepL, and ChatGPT, AI is transforming how people access and understand
information across linguistic barriers. These technologies promise speed, cost-effectiveness, and
accessibility, offering users an unprecedented ability to engage with content in multiple
languages.
However, while AI translation systems have made remarkable technical progress, concerns
remain about their ability to capture linguistic nuance, cultural context, and the subtleties of
human expression. For instance, idiomatic phrases, tone, and regional variations often present
significant challenges to automated systems. Furthermore, issues of trust, perceived accuracy,
and bias in AI-generated translations raise questions about their reliability in sensitive or
professional contexts, such as legal, medical, or diplomatic communication.
Although much of the existing research in this area has focused on quantitative assessments of
AI translation quality – such as BLEU scores or error rates – there is a growing need to explore
the human dimension: how individuals experience and perceive AI translation tools in real-world
contexts. In particular, the voices of professional translators and frequent users of such tools are
underrepresented in academic literature.
Methods
This study employs a qualitative research method to examine how AI-based translation tools are
experienced and interpreted by both professional translators and end-users. The STS framework
guides the exploration of how technology (AI translation systems) and human elements (users,
workflows, expectations) interact, adapt, and shape each other within translation contexts.
Rather than evaluating AI translation tools on their linguistic accuracy alone, this research aims
to capture the social, experiential, and interpretive dimensions of their use. This is done through
in-depth engagement with participants to understand the perceived benefits, limitations, and
transformations in translation practice caused by AI integration.
Volume 15 Issue 06, June 2025
Impact factor: 2019: 4.679 2020: 5.015 2021: 5.436, 2022: 5.242, 2023:
6.995, 2024 7.75
http://www.internationaljournal.co.in/index.php/jasass
103
Results
Analysis of the data through a sociotechnical lens revealed five central themes reflecting the
evolving relationship between humans and AI in translation:
efficiency vs. accuracy, post-editing
demands, cultural sensitivity, trust and bias,
and
changing translator roles.
These themes
illuminate the broader structural and cognitive adjustments taking place as translation becomes
increasingly mediated by AI technologies.
Efficiency vs. Accuracy.
AI translation systems are widely integrated into personal and
professional routines due to their unparalleled efficiency. Users often prioritize speed and
accessibility, particularly when the translation task does not require deep semantic or contextual
accuracy. In these cases, AI is seen as a practical tool that facilitates immediate comprehension
across language barriers.
However, this efficiency frequently comes at the cost of precision. AI-generated translations can
exhibit surface fluency while failing to convey deeper linguistic nuance or contextual meaning.
This discrepancy reflects a core tension within sociotechnical systems: the trade-off between
automation and interpretive quality. As users become more reliant on AI, their expectations of
translation begin to shift toward utilitarian goals, potentially narrowing the perceived value of
linguistic depth and cultural richness.
Post-editing Needs.
The prevalence of post-editing in AI-assisted translation workflows reveals
how automation does not eliminate human labor but rather transforms its nature. Instead of
producing translations from scratch, professionals are increasingly engaged in tasks of revision,
correction, and semantic alignment. This shift introduces a new form of cognitive labor that is
often invisible but essential for ensuring the accuracy and appropriateness of AI outputs.
Post-editing requires not only linguistic proficiency but also critical evaluation skills, as
practitioners must discern whether the AI has preserved meaning, tone, and intention. In this
hybrid model of translation, the human role becomes one of interpretive oversight, positioned
downstream from the machine’s initial output. This change reflects a broader restructuring of
authority within the sociotechnical system, wherein the machine initiates linguistic content and
the human becomes the final arbiter of its quality.
Cultural Sensitivity.
Despite advances in natural language processing, AI translation systems
continue to exhibit significant limitations in handling culturally embedded content. Idioms,
metaphors, humor, and region-specific references are frequently mistranslated or rendered
inappropriately, highlighting the system's inability to fully grasp context-dependent meanings.
These failures underscore a fundamental limitation of current AI: its reliance on pattern
recognition rather than embodied cultural understanding.
Trust and Bias.
Trust in AI-generated translations is contingent on context, familiarity, and
perceived risk. Users tend to place greater trust in AI for informal or low-stakes tasks but express
hesitation in relying on it for sensitive or high-impact communication. This conditional trust
points to an awareness of the systemic opacity of AI systems – users recognize that they have
limited insight into how translation decisions are made and what influences them.
Changing Roles of Translators.
AI adoption is catalyzing a redefinition of professional
translation roles. Rather than serving solely as linguistic producers, translators are increasingly
positioned as quality controllers, editors, and consultants. This evolution reflects a redistribution
of labor within the sociotechnical ecosystem, where machines handle the mechanical aspects of
translation and humans focus on higher-order interpretive tasks.
Volume 15 Issue 06, June 2025
Impact factor: 2019: 4.679 2020: 5.015 2021: 5.436, 2022: 5.242, 2023:
6.995, 2024 7.75
http://www.internationaljournal.co.in/index.php/jasass
104
Discussion
This article explored how artificial intelligence is reshaping translation practice, using
Sociotechnical Systems Theory to interpret the interplay between human users and AI tools. The
findings highlight a complex dynamic: while AI offers speed and accessibility, it also introduces
new challenges around accuracy, cultural sensitivity, and professional identity.
AI has become a key actor in translation workflows, not simply automating tasks but redefining
roles. Translators are now more often editors and consultants, overseeing the refinement of
machine-generated content. This reflects a shift in how expertise is understood – away from
production and toward judgment, quality control, and cultural mediation.
However, this shift comes with tensions. Users often accept surface-level accuracy for
convenience, but professional settings still demand high-quality, culturally appropriate
translations. AI’s difficulty with idioms, tone, and bias underscores its limitations, particularly
when representing diverse voices or politically sensitive material.
Ethically, the growing reliance on AI risks undervaluing human labor and reinforcing
algorithmic biases. Transparency in how AI systems are trained, and clearer boundaries around
appropriate use, are essential. For translation to remain both efficient and meaningful, human
oversight must remain central. Overall, AI is transforming – not replacing – human translation,
demanding new skills, ethical frameworks, and a rethinking of what quality means in
multilingual communication.
Conclusion.
AI is rapidly transforming the field of translation, offering greater speed and access
while simultaneously challenging traditional notions of quality, authorship, and cultural nuance.
This study, through a sociotechnical lens, shows that while AI enhances efficiency, it cannot
fully replace human insight – especially in contexts requiring deep linguistic judgment and
cultural sensitivity.
As translators adapt into roles as editors and consultants, the profession is evolving rather than
disappearing. However, to ensure responsible and equitable use of AI in translation, there is a
growing need for ethical oversight, inclusive design, and sustained human involvement. Future
developments in AI must focus not only on technical accuracy but also on the preservation of
meaning, diversity, and trust in cross-cultural communication.
References
1.
Bowker, L. (2002).
Computer-aided translation technology: A practical introduction
.
University of Ottawa Press.
2.
Hutchins, W. J. (2005).
The history of machine translation in a nutshell
.
3.
Köhler, M., & O’Brien, S. (2018). Post-editing practices: A survey.
The Journal of
Specialised Translation
, 29, 4–23.
4.
Latour, B. (2005).
Reassembling the social: An introduction to actor-network-theory
.
Oxford University Press.
5.
Nababan, M. R., Hasyim, A., & Gunawan, B. (2020). Ethical considerations in machine
translation.
Journal of Translation Studies
, 15(2), 102-119.
6.
Sundaram, H., & Sinha, A. (2021). Bias and fairness in AI-powered language translation.
AI & Society
, 36(3), 683–692.
7.
Suchman, L. A. (2007).
Human-machine reconfigurations: Plans and situated actions
(2nd ed.). Cambridge University Press.
Volume 15 Issue 06, June 2025
Impact factor: 2019: 4.679 2020: 5.015 2021: 5.436, 2022: 5.242, 2023:
6.995, 2024 7.75
http://www.internationaljournal.co.in/index.php/jasass
105
8.
Vasconcellos, M. (2019). Translator’s role in post-editing machine translation: A
sociotechnical perspective.
Translation Spaces
, 8(1), 41-59.
9.
Wang, Y., & Li, X. (2023). Cultural challenges in AI translation: A qualitative study.
Language and Intercultural Communication
, 23(1), 56-72.
