In February 2022, while translating Dana Spiotta's novel Wayward into French, literary translator Yoann Gentric tested whether AI could replace him. He fed the sentence "Bright, sharp night air, bracing" into DeepL, a machine translation engine. The AI rendered it as L'air de la nuit, vif et vif, était vivifiant, missing the absurd repetition. His own translation, L'air pur et piquant de la nuit, vivifiant, was far superior. However, when he repeated the test in spring 2024, DeepL produced L'air nocturne était vif, pur et vivifiant, showing improvement with varied vocabulary, albeit with an added verb.
AI chatbots powered by large language models (LLMs) are rapidly infiltrating the translation industry in Europe, home to over 200 languages and a booming tech sector. According to a joint survey by French authors' societies ADAGP and SGDL, 79% of translators believe AI threatens to replace all or part of their work. In Britain, a 2025 survey found 84% expect lower demand and pay.
For many, change is already here. Laura Radosh, a Berlin-based German-to-English translator, saw job requests drop from four per month to one, with most being "post-editing" tasks—correcting machine-translated texts. "Post-editing took me as much time as translating from scratch," she said. It is also less creatively fulfilling and poorly paid, often by the hour instead of by the page. German publishers offer two to eight euros per page for post-editing, a quarter of standard rates.
Even before LLMs, translation was precarious. The German translators association VdÜ found literary translators earn an average of €20,363 per year before tax. Radosh recently took a part-time bookkeeping job. Marco Trombetti, CEO of Translated, notes that human translators produce about 3,000 words daily, but AI changes the economics.
AI's Limitations and Human Strengths
Machine translation still struggles with context. In 2024, Springer Nature's auto-translation rendered "capital" as Hauptstadt (capital city) instead of Kapital in a book titled 'Capital' in the East: Reflections on Marx. Springer Nature said the error was rare and the pilot has ended.
Jörn Cambreleng of Atlas emphasizes that machine translation lacks creativity, producing generic sentences. Human translators strive for originality. Katy Derbyshire, a Berlin-based translator, notes that AI cannot handle dialogue well. "My body has experienced pain and joy that literature conveys. I understand what someone might scream when they hit their toe—an algorithm doesn't."
Literary translation appears safer than technical translation. Harlequin France uses AI for pulp novels but post-edits them. In Germany, translated literature made up 15% of new books in 2024, a historic high. Authors increasingly contractually ban AI use in translation, says Marieke Heimburger of VdÜ.
Fernando Prieto Ramos of the University of Geneva reports a drop in translation course applications three years ago, but the trend is reversing with diversified training. Even AI developers concede limitations. Trombetti notes that translating "Solo tre parole: non sei solo" into English yields "Just three words: you are not alone"—four words, not three, which AI struggles with.
Heimburger concludes: "I am not scared of AI because I know it cannot do what I can do. What I am afraid of is the people who think that AI can do my job."
