What is Ethical in Translation with AI?

Ethics in Artificial Intelligence is a popular debate in our times. Professionals in the translation field and language services often question themselves about “what is the right thing to do.”

What is ethical and unethical in translation with AI integration? Does the use of AI represent a less ethical stance or are there nuances to consider?

Ethics: from standard philosophy to the postmodern era

Let’s start understanding ethics. This word is derived from the Greek “éthos”, and it is most commonly used to refer to what philosophy calls studies of morality. It helps society to define what is right or wrong and judge the propriety of character. 

In the translation industry, ethics protects authors, clients, translators, and end-users. It’s a shared responsibility between all the participants in this relationship, but we focus here on the view of translators and translation industries. I.e., the service provider.

Examples of unethical conduct ‍

It’s always better to understand if you paint the images in your head. So let’s consider the examples and the following situations. 

1. Mistranslation 

The most complex form of unethical posture inside translation is mistranslation. It can include many sides of the same coin, such as adding or omitting content.

This can happen by purpose or due to the translator’s lack of ability. Both examples are unethical. 

Suppose you take on a project without having the right qualifications to do so. In that case, there is a probability of omitting information because you don’t get the proper meaning, or you don’t have the discernment to identify what is important, or you misinterpreted the arguments. 

There is a high chance that the result is an undesirable change of the original message, right?

Mistranslation by purpose, following a second intention, is even worse. The causes may rely on the translator's interpretations. Suppose a case where the translator wants to impose their point of view in the content, for aesthetic or political reasons. 

Another example is a misunderstanding about what authorship means. Having authorship is not a problem at all, but changing the original text to the point you cannot capture the style and main ideas of the original author anymore, is an unethical posture. 

2. Plagiarism

Plagiarism is well-known in the world of digital content. It’s pretty hard to identify the first author in this chaos. 

In the translation field, plagiarism can be using someone else's translation without having permission (monolingual plagiarism) or translating content from another language and taking credit for it (cross-language). 

If you are hired to work on a project to translate a book that already has one or more versions translated by another professional… Well, you can even consult the first one (if you are strong enough not to get influenced), but never copy and paste the job that is already done.

In fact, why do that? Much better if you create your version, seasoning it with your style of work. 

Using generative AI with responsibility can make the workflow easier and you won’t need to use plagiarism as a saving time strategy.

3. Break Confidentiality

Many translators and translation agencies work with confidential projects. Legal translations, science translations, business documents, academic interviews with valuable testimonials, and many others that require secrecy. 

Disclosing confidential information is not only unethical but also a trust relationship breaker. You can miss the chance to get another service contract with the same client.

Besides, depending on how valuable was the information, it can lead to lawsuits.

Better to stay out of trouble, agreed?

4. Inappropriate Budgets

Ethics also shows in the deal moments. Selling a translation service speaking in terms of business, but also respecting your time and work is a trick. You have to find a balance, a deal that is good for both sides and is coherent with the complexity of the project.

Sometimes, especially if you are beginning in this field, you can make a mistake with the budget. 

Imagine you propose to your client an initial budget considering your hours of work, or the number of words, but you finally learn how to use AI management systems to improve your workflow.

So now you are translating twice as fast. Much less work, time-saving, equal or best performance. Is that fair to keep the same budget and pocket the excedent? Another ethical question is on the table.

Maybe, a strategic exit is to adjust your client's budget, offering him better cost-benefits and the possibility to take on more projects (since you have more time and efficiency). More projects with the same client or with another one. Seems good, right? Probably the clients will be happy with the offering.

At the same point, your budget mistake could have been to cover less than the project requires. It’s also an ethical posture to readjust and cover more! Besides, it’s a self-respecting decision with your efforts and investments. 

The professor of Translation Studies and head of the Languages unit at Tampere University, Finland, Kaisa Koskinen, in her article “Beyond Ambivalence: postmodern and the Ethics of Translation” remembers that a translator may face several dilemmas on duty. 

There is not a ready-made list indicating the best decisions. For her, it is an ambivalent discussion: “Ethics can never be resolved once and for all since moral aspects need to be continuously renegotiated and conditions and situations change.”

In our post-modern society, ethics needs to attend to the questions of our times. Today we talk about augmented translation, with technology, postcolonial translators, and even feminist translators. It’s not about the subject in matters, but the approach we give to the content, it is also a political agenda. 

All the basic aspects we cite before are important, but also overcoming biases, negotiating with the original content, and compromising with the end-reader and consumers. 

Ethics in the translation with AI integration isn’t different from ethics in the translation field in general. Unethical conduct also could have been practiced without AI. The thing is: a good well-intentioned professional will find the tools to make it right, and the same serves the opposite.

References:

KOSKINE, Kaisa. Beyond Ambivalence: postmodern and the Ethics of Translation. University of Tempere. Tempere, 2000. Available at: 

https://trepo.tuni.fi/bitstream/handle/10024/67049/951-44-4941-X.pdf?sequence=1&isAllowed=y. Access on 19 of June of 2024.