Translation in the Age of AI Agents
الترجمة في عصر الذكاء الاصطناعي
DOI:
https://doi.org/10.33705/1111-019-001-001Keywords:
AI agents, machine translation, large language models, agentic translation, domain grounding, human-in-the-loop, ProofAgent HarnessAbstract
Translation is entering a new technological phase. For decades, machine translation systems converted sentences from one language to another with increasing fluency, yet they remained constrained by context loss, domain ambiguity, cultural nuance, and limited mechanisms for verification. The emergence of large language models (LLMs) and AI agents changes the translation function from a passive text-conversion task into an active, tool-using, context-aware workflow. An AI translation agent can retrieve domain knowledge, consult glossaries, check terminology, preserve institutional tone, compare alternative translations, evaluate uncertainty, and request human review when risk is high. This paper argues that agentic translation represents a paradigm shift in diplomatic, scientific, technical, legal, healthcare, and enterprise communication. The shift is especially important because the web is linguistically imbalanced: English dominates online content, while French and Arabic occupy much smaller shares, shaping the available knowledge base for AI systems. This imbalance creates both an opportunity and a risk. Agentic translation can bridge knowledge gaps by grounding translation in trusted corpora and domain-specific references, but it must be evaluated through human-in-the-loop processes and automated adversarial evaluation frameworks such as ProofAgent Harness. The paper proposes a reference architecture for agentic translation, identifies key domain use cases, and outlines evaluation dimensions for reliable deployment.
References
[1] Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., and Polosukhin, I. (2017). Attention Is All You Need. Advances in Neural Information Processing Systems. https://arxiv.org/abs/1706.03762
[2] Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., et al. (2020). Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems. https://arxiv.org/abs/2005.14165
[3] Meta AI. (2022). No Language Left Behind: Open-source models for translation across 200 languages. https://ai.meta.com/research/no-language-left-behind/
[4] Bapna, A., Caswell, I., Kreutzer, J., Firat, O., van Esch, D., Siddhant, A., Niu, M., Baljekar, P., et al. (2022). Building Machine Translation Systems for the Next Thousand Languages. arXiv:2205.03983. https://arxiv.org/abs/2205.03983
[5] Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., et al. (2020). Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. Advances in Neural Information Processing Systems. https://arxiv.org/abs/2005.11401
[6] Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., and Cao, Y. (2022). ReAct: Synergizing Reasoning and Acting in Language Models. arXiv:2210.03629. https://arxiv.org/abs/2210.03629
[7] Schick, T., Dwivedi-Yu, J., Dessì, R., Raileanu, R., Lomeli, M., Zettlemoyer, L., Cancedda, N., and Scialom, T. (2023). Toolformer: Language Models Can Teach Themselves to Use Tools. Advances in Neural Information Processing Systems. https://arxiv.org/abs/2302.04761
[8] Bousetouane, F. (2025). Agentic Systems: A Guide to Transforming Industries with Vertical AI Agents. arXiv:2501.00881. https://arxiv.org/abs/2501.00881
[9] Singh, A., Ehtesham, A., Kumar, S., Khoei, T. T., and Vasilakos, A. V. (2025). Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG. arXiv:2501.09136. https://arxiv.org/abs/2501.09136
[10] W3Techs. (2026). Usage statistics of content languages for websites. Accessed June 13, 2026. https://w3techs.com/technologies/overview/content_language
[11] Taguchi, C., Mai, S., Kurabe, K., Sakai, Y., Agyei, G., Eslami, S., and Chiang, D. (2025). Languages Still Left Behind: Toward a Better Multilingual Machine Translation Benchmark. arXiv:2508.20511. https://arxiv.org/abs/2508.20511
[12] Bousetouane, F. (2026). ProofAgent Harness: Open Infrastructure for Adversarial Evaluation of AI Agents. arXiv:2605.24134. https://arxiv.org/abs/2605.24134
[13] ProofAgent. (2026). ProofAgent Harness: Open-source infrastructure for adversarial AI agent testing. https://www.proofagent.ai/harness
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 فؤاد بوسطوان
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
