FIT publishes report on the ethics of AI in translation and interpreting

The International Federation of Translators (FIT) has published a new report, Ethics of artificial intelligence in translation and interpreting, which sets out the principles that should guide the responsible use of AI in the profession. The report was shared with FIT's member associations at the Federation's annual general meeting, attended by our CEO Sara Robertson on 19 September.

About FIT

FIT is the international federation of professional associations of translators, interpreters and terminologists. It was established in 1953 under the auspices of UNESCO and today brings together more than 100 member associations around the world. Its aim is to promote professionalism, improve conditions for the profession and uphold translators' rights and freedom of expression.

FIT members elect the Council at a Statutory Congress held every three years, most recently in Geneva in September 2025. The Federation has now moved to holding an annual general meeting online in the years between congresses, to improve communication with members and allow it to progress projects more quickly. This year's meeting was the first in the new online format.

About the report

The report was prepared by the FIT Working Group on AI Ethics, chaired by Adriana Zúñiga Hernández of ACOTIP (Costa Rica), with members drawn from associations in Europe and the Americas. Ginta Rudd and Irina Sanders from the Association of Interpreters and Translators (AIT) contributed the UK perspective.

The report identifies six interrelated ethical principles for the use of AI in translation and interpreting:

• accessibility
• accountability
• accuracy
• confidentiality
• fairness
• transparency

The report makes clear that FIT does not seek to discourage innovation. Instead, it recognises that AI can support translation and interpreting workflows, but concludes that AI cannot replace the professional judgement, ethical responsibility and cultural competence of qualified translators and interpreters. Human oversight remains essential, particularly in high-stakes settings such as law, healthcare, education, finance and diplomacy.

Among its recommendations, the report calls for responsibility for AI errors to be shared between developers, organisations and the professionals who use AI tools, rather than resting with end users alone. It also argues for clear labelling so that people know whether a translation or interpreting was produced by a human, by a machine, or by a machine and then reviewed by a professional. In addition the report highlights the role of professional associations in promoting ethical practice, keeping members informed and contributing to the development of regulation.

What the report means for members

The report's findings reflect our own position that human expertise remains central to high-quality translation and interpreting, and that AI should support qualified professionals rather than replace them.

We will continue to work with FIT and partner associations, in the UK and internationally, to help members understand AI and how to use it responsibly (if they choose) and to press for appropriate safeguards.

The full report is available to download from our Knowledge Hub.