In recent weeks, the use of artificial intelligence in literary translation has sparked strong reactions. The concerns expressed by many translators are legitimate and deserve to be heard. But they should not obscure the essential point: too often, the current debate relies on false oppositions, when the reality is more complex, more nuanced and far more urgent. The question is not whether one should be “for” or “against” artificial intelligence. The real question is how it is used, by whom, and under what responsibility.
1. A transformation already underway
Translation is one of the first intellectual professions to be deeply impacted by AI. For several years, companies and institutions have increasingly relied on automated translation tools used directly, often without communication, transparency or systematic human oversight. This silent shift has already led to a significant reduction in work volumes entrusted to professional translators. This transformation is not the result of isolated editorial decisions. It stems from the widespread, and often uncoordinated, adoption of automated tools used without qualified human intervention. This is where the real issue lies.
2. Raw automation or responsible innovation
Equating any use of AI with an intention to replace translators is a mistake. There is a fundamental difference between: • raw automation, where texts are produced and delivered without human accountability; • and hybrid models, in which technology is used as an assistance tool, under the continuous supervision of qualified professionals. In these hybrid models, AI does not decide it suggests. Human translators remain fully responsible for the final text, including meaning, tone, intent and emotional depth. Ignoring this distinction risks leaving the field open to the most dehumanised uses of technology.
3. Human responsibility as the dividing line
At Fluent Planet, we have made a clear choice: no text is delivered without qualified human intervention, and translators always retain responsibility for the final output. Our translation assistance tools are designed to support linguists in their work, strengthen consistency and improve efficiency never to replace human expertise or editorial judgement. This choice is neither technophile nor conservative. It is pragmatic.
4. Protecting quality, employment and reader trust
Readers, across all publishing genres, are highly sensitive to the quality of the texts they read. The effects of mechanical translation are immediately noticeable: stylistic flattening, cultural awkwardness, and loss of emotional intensity. Opposing literary quality and technological innovation is a false dilemma. What truly threatens quality, professional livelihoods and public trust is the abandonment of human responsibility in favour of ungoverned automated solutions.
5. Raising the level of the debate
Artificial intelligence is neither an absolute evil nor a miracle solution. It is a powerful tool that requires clear governance and responsible choices. The real debate is therefore not: “Will AI replace translators?” but rather: “What kind of model do we want to build so that technology serves quality, culture and skilled employment?” The translation profession has always evolved alongside its tools. What is new today is the scale of change and therefore the collective responsibility that comes with it. A call for collective responsibility Publishers, language service providers, translators, institutions and technology actors all share a common responsibility: not to confuse innovation with abdication. Rejecting technological evolution altogether leaves the door open to the most brutal forms of automation. Integrating technology without a framework risks permanently weakening a profession. The only sustainable path forward is one of responsible, governed and deeply human innovation. That is the path Fluent Planet has chosen and will continue to pursue.
Fluent Planet A French language services company committed to quality, responsibility and human-centred translation in the age of artificial intelligence.
