Applied AI at work / Module 8
Module 8. Drafting and translating in English, French and German
Time: 3 h (lesson 50 min, exercise 1 h 50 min, quiz 20 min) For: anyone who writes to clients, customers, candidates, colleagues or authorities in more than one language: HR, sales, operations, legal, client service, fund administration, audit and finance.
Work in Luxembourg happens in several languages at once. In STATEC's 2018 labour force survey, three resident workers in four work in a multilingual environment and French is the language used by the largest number at work (78%) (STATEC, Regards 09/2019: https://statistiques.public.lu/fr/publications/series/regards/2019/regards-09-19.html). Translation is also one of the top five AI use cases the CSSF found in the sector, at 19% of reported use cases (CSSF thematic review, May 2025, p. 6: https://www.cssf.lu/wp-content/uploads/Thematic_report_AI_2024.pdf).
The Commission addresses this case directly. Asked whether a company whose employees use ChatGPT for "writing advertisement text or translating text" must comply with the AI literacy requirement, its answer is: "Yes, they should be informed about the specific risks, for example hallucination" (AI Literacy Q&A: https://digital-strategy.ec.europa.eu/en/faqs/ai-literacy-questions-answers; Article 4 of the AI Act: https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-4). This module is about those risks in practice: a translation that reads well and says something else.
What you will be able to do
- Prepare a text for translation with placeholders so that no investor or client data reaches a public tool.
- Write a translation brief that fixes glossary terms, register, number and date formats, and forbids additions and omissions.
- Review an AI translation side by side and classify errors: number or date, negation, defined term, entity or role, register, false friend.
- Use correct EN, FR and DE fund terminology for NAV, management company, depositary, subscription, redemption and cut-off time.
- Draft English investor communications from a fact list and prove that every statement traces back to a fact.
3 video lessons, 11 minutes
- 8.1 A translation tool is a third party, and where it fails 4:02
- 8.2 Translate with a brief, then check like a reviewer 3:17
- 8.3 Drafting in English, what never to do, and the exercise 3:48
Exercise: Module 8 exercise: review an AI translation of a notice to shareholders
Fictional data, scored automatically against the answer key, with the model answer after you submit.
Module quiz
8 questions, pass mark 6. Passing it counts towards your course certificate.
Sources
- STATEC, Regards 09/2019, languages at work (labour force survey 2018): https://statistiques.public.lu/fr/publications/series/regards/2019/regards-09-19.html
- CSSF, Thematic review on the use of artificial intelligence in the Luxembourg financial sector, May 2025, p. 6: https://www.cssf.lu/wp-content/uploads/Thematic_report_AI_2024.pdf
- European Commission, AI Literacy, Questions and Answers: https://digital-strategy.ec.europa.eu/en/faqs/ai-literacy-questions-answers
- Regulation (EU) 2024/1689 (AI Act), Article 4, as amended, AI Act Service Desk: https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-4