By Mahmoud Kamel
Artificial intelligence is changing translation education, not only because it can translate texts, but also because it is changing the way students learn, instructors assess their work and universities understand academic honesty. With tools such as ChatGPT becoming increasingly capable of producing fluent translations, an important question comes to mind: How can we prepare future translators to use AI effectively without losing the skills, judgment and responsibility that make them professionals?
Not every student has the same access to AI
Another important issue discussed in the chapter is digital inequality. It is easy to assume that all students have equal opportunities to use modern technology, but this is not always the case.
Some students can afford paid subscriptions to advanced AI and translation tools, while others depend on free versions with fewer features. Some have reliable internet connections and suitable computers, while others face financial or technical difficulties. Access may also be affected by geographical restrictions and local regulations.
These differences can influence academic performance. If an assignment allows students to use advanced AI tools, those who have access to better technology may enjoy an advantage over classmates who cannot afford it. The result may be an assessment that measures not only translation competence, but also a student’s financial resources and technological opportunities.
This raises a question of fairness. How can a university compare the performance of students who have worked with different levels of technological support?
One possible response is to ensure that students have equal access to the tools required for a particular assignment. Another is to design assessment tasks that make the student’s own knowledge and decision-making visible, rather than allowing the quality of the final text to depend mainly on the technology available.
Fairness in the AI era does not mean giving every student exactly the same tool in every situation. It means making sure that differences in access do not unfairly determine educational results.
Can instructors know whether a student used AI?
Transparency is another major challenge. Universities may require students to disclose whether they have used AI, but such rules are not always easy to enforce. Some students may be reluctant to admit using these tools, even when their use is permitted under certain conditions.
Instructors may turn to AI-detection software in an attempt to identify machine-generated work. However, the chapter discusses the limitations of relying on detection and the risk that decisions may be influenced by suspicion rather than reliable evidence. A student’s writing may be wrongly questioned, while the use of AI in another assignment may go unnoticed.
This creates a difficult situation for both students and instructors. If the institution depends too heavily on detection, assessment can become a contest in which students try to conceal their use of AI and instructors try to discover it. Such an approach does little to encourage honest learning.
A more constructive solution is to make the translation process part of the assessment. Students can be asked to explain the difficulties they encountered, identify the changes they made to an AI-generated draft and justify their translation choices. They may also be required to provide the prompts they used, the machine’s initial output and a record of their revisions.
These measures would help instructors understand not only what the student submitted, but also how the final translation was produced. They would encourage students to take responsibility for their decisions and make their own contribution clearer.
The purpose should not be to catch students out. It should be to create an educational environment in which honest use of technology is possible and the student’s real learning can be assessed.
When AI reproduces human bias
The ethical questions about AI are not only about academic honesty. They are also about the language, values and ideas that AI systems may repeat in their translations.
AI systems learn from large amounts of existing texts and other data. If these materials contain stereotypes or unfair views, AI may repeat them in its translations. This can affect how people, cultures, men and women, and different social groups are presented in another language.
For example, a translation may use words that support a stereotype or present a certain group in an unfair way. This problem may not be easy to notice, especially when the translation is grammatically correct and sounds natural.
This is especially important when translating sensitive texts about identity, politics, culture or social relations. A translator must not only check whether the words are correct, but also whether the translation gives the meaning of the original without adding unfair or harmful ideas.
The problem becomes more difficult when we cannot clearly understand how an AI system reaches its results. If an AI translation contains biased or discriminatory language, who is responsible? Is it the developers who designed the system, the institution that uses it or the translator who accepts its output?
More than one party may be involved, but this does not remove the translator’s responsibility to check the final text. The chapter, therefore, stresses the importance of teaching students how to recognise bias, question the ideas behind AI-generated language and understand the possible results of their translation choices.
Teaching students to think, not just use technology
One important idea in the chapter is that translation education should not only teach students which tools to use and how to use them. Students also need to understand when a tool is suitable, when its results should be questioned and what may happen if they accept its suggestions without checking them.
This requires what the authors call digital reflexivity. In simple words, it means thinking carefully about how we use technology and examining our own decisions when working with it.
A student should not accept an AI translation simply because it sounds good. The student should compare it with the original text, check important information and terminology, and consider whether the machine has understood the context correctly. Students should also know that AI can give answers that sound convincing but may be wrong or have no clear support.
Instructors can develop these skills through practical classroom activities. They can ask students to compare different AI translations, find possible bias, identify mistakes and explain why they accepted or rejected certain suggestions. Students can also write short comments explaining their choices and the reasons behind them.
These activities would make ethics a practical part of translation training instead of a subject discussed only in theory. They would also help students understand that professional translation is not only about producing a fluent text. It also requires good judgment, an understanding of the reader and responsibility for the meaning of the message.
The difference between academic and professional ethics
The chapter also discusses the relationship between ethical behaviour at university and ethical behaviour in professional translation. The two are closely connected, but they are not always the same.
At university, students may be asked to explain when and how they use AI so that instructors can assess their work fairly. In professional translation, however, the rules for using translation technology may depend on the agreement with the client, the type of work and the need to protect confidential information.
A professional translator may use a dictionary or a translation tool without mentioning every resource used. But AI can raise additional questions, especially when the work contains confidential documents or sensitive information. The translator must consider whether the tool is suitable and whether sending information to an external system could put the client’s data at risk.
Students, therefore, need to understand the difference between using AI as a learning tool and using it to provide a professional service. They should also learn how to explain their decisions when questions arise about the quality, accuracy or ethical side of their work.
This is an important responsibility for universities. They should prepare students not only to pass their examinations, but also to deal with the real ethical problems they may face after graduation.
Rethinking the way translation is assessed
Perhaps the most important point in the chapter is that universities need to reconsider how they assess translation skills in the age of AI.
If instructors look only at the final translation, it may become more difficult to know what a student can really do. A well-written text does not always mean that the student understands the original, can identify mistakes or knows how to make the right translation decisions.
The authors suggest giving more attention to the process of translation, not only to the final result. Instead of asking for a completed translation alone, instructors can ask students to record the steps they followed, explain how they used AI and give reasons for their changes. Short learning journals, translation comments and self-assessment activities can help instructors understand how students reached their final results.
Universities can also combine assignments that allow the use of AI with classroom exercises or examinations where students must show their skills without it. In this way, instructors can assess both traditional translation skills and the ability to use modern technology critically.
These changes would make assessment closer to the needs of professional translation. In the translation industry, knowing how to use AI is becoming more useful, but knowing how to check its results is still essential.
The aim is not to return to a world without technology. It is to make sure that technology helps students learn instead of doing the learning for them.
A more responsible future for translation education
The discussion by Bo Li and Yuan Liu reminds us that AI is developing faster than many educational policies and ethical rules. Universities, instructors and students are still learning how to deal with tools that can perform tasks once seen mainly as human language skills.
There is no single rule that can solve every problem. A rule suitable for a simple classroom exercise may not be suitable for a final-year research project. In the same way, the rules for using AI in a translation examination may be different from those for professional work for a client.
What is needed is a clear framework that explains how AI can be used, protects academic honesty, deals with differences in access to technology and encourages students to think carefully about the tools they use. Ethical questions should be discussed openly, and students should have opportunities to practise making responsible decisions before they start their professional careers.
Translation education must also recognize that technology does not remove the need for human responsibility. In fact, as AI becomes better at producing translations that sound convincing, critical thinking becomes even more important. Students need to learn not only how to get an answer from a machine, but also how to decide whether that answer can be trusted.
The responsibility remains human
Chapter 10 presents an important view of the future of translation education. The main question is not simply whether AI should be allowed in the classroom, but how universities can prepare students to use it honestly, fairly and responsibly.
AI can support translation practice, help students consider different ways of translating and provide useful help during learning. However, it can also create problems related to academic honesty, unequal access, hidden bias and questions about who is responsible for the final text.
The answer is neither to accept every new technology without thinking nor to reject it because of fear. The better way is to develop clear rules, effective teaching methods and a stronger sense of professional responsibility.
For translation instructors, this means helping students become more than good users of AI. They should become critical thinkers who understand the limits of technology, can explain their decisions and take responsibility for their work.
Future translators will need technological knowledge, language skills and cultural awareness. But these qualities will not be enough without ethical judgment.
AI may change the way a translation is produced. It should not change our understanding of honesty, fairness and responsibility. These principles must remain at the heart of translation education, even as the tools continue to develop.
Dr Mahmoud Kamel
Professor at the Academy of Arts
