By Nataliya Horbachevska, CEO, Task Force
In the translation industry, AI has been a commercial reality for over a decade. While the recent hype surrounding ChatGPT and similar models has drawn widespread attention, for professional language service providers (LSPs), AI has long been embedded in core operations. Machine translation, AI-enhanced CAT tools, and automated post-editing have been used daily to serve clients efficiently, sometimes complemented by human oversight, sometimes fully automated. The recent surge of interest is less a revolution and more a recognition of what many translation teams have already been doing strategically for years.
Mirko Plitt, a pioneering computational linguist with decades of experience developing and implementing machine translation solutions, says: “The language service industry is one of the sectors with the longest practical AI experience, even if it may not yet have been labelled AI at the time. For this industry, automation of human-intelligence activities hasn’t been a sudden discovery when ChatGPT went public, but a gradual process that has spanned decades. A look back onto how the language industry has been evolving can therefore offer other industries some valuable insights into approaches to transformation that have been successful or not”.
Dispelling the myth: AI will replace translators
Amid the AI buzz, there’s been a recurring narrative: “AI will replace human translators.” Yet for those of us in the industry, this is hardly new. Commercial use of AI in translation predates the current hype—Google Translate and other automated tools have existed for years, and yet professional translation services remain essential. The reality is that AI is a tool, not a replacement. Its usefulness depends on task complexity and the client’s goals. Just as Excel did not make accountants obsolete, and Thermomix or pressure cookers did not empty restaurants, AI does not eliminate the need for human expertise. Complex content, context-sensitive decisions, and nuanced communication still require skilled linguists.
The evolving nature of innovation: shifting roles, risks, and the human factor
“Innovation in language service industry has mainly been driven by large service buyers, not service providers, and among the latter even less so the large service providers. Among the large service buyers, innovation has often been opportunistic, not strategic. For most established businesses, disruptive innovation is risky and therefore bad strategy”, says Mr Plitt.
Indeed, while innovation may have been instigated by seasoned individuals, for the innovation to spread, new generations had to arrive to the job market, both at the service providers and buyers. An important vector for the spread of innovation is also down to individuals moving between employers, which follows a natural pace. What really matters (regardless of the degree of possible automation) is business expertise – a deep understanding of internal and external client needs.
Contrary to the expectation set by the hype surrounding AI, the picture that emerges in language services isn’t one of sudden disruption but of carefully managed transition towards new models being shaped at their edges by loose networks of creative, typically rather experienced, individuals with sufficient leeway to innovate, and with access to a workforce ready to adopt new ways of working.
What are new ways to use AI in language service industry?
One of the most prominent tools in the field of AI and translation today is computer-assisted translation (CAT) tools. CAT tools have long been a staple in the translation industry. They allow us to store translation memory, manage glossaries, and maintain quality control over translations. These tools are now being integrated with AI features, such as prompts for segment style adjustments, terminology replacement and even automated quality checks.
For example, AI-enhanced CAT tools can now automatically suggest stylistic changes for certain segments, replace terminology according to predefined rules, and execute quality control tasks. While these AI capabilities add significant value, they require a high level of technical expertise to integrate and use effectively. This is where professional language service providers (LSPs), like ours, with both linguistic and technical expertise, come into play.
Another best practice of today is to apply machine translation and then use AI to perform a post-editing check. With this approach, machine translation handles the bulk of the work, and AI is used to review and improve the quality. AI can rank segments based on translation quality, and only those with lower scores are sent to human editors for final review. This reduces the manual effort and saves time and costs. However, implementing such solutions requires significant investment in technology and integration, which may not always be cost-effective compared to traditional human translation, and hence – demanded only by the clients with huge volumes of multilingual content.
Risks of using AI without human oversight
Despite the clear benefits, relying solely on AI in translation carries significant risks, such as hallucinations, lack of clarity regarding confidentiality of the content you want to translate, and often – loss of nuance and context.
Krystyna Konovalova, language service coordinator of the World Council of Churches in Geneva, says: “As the coordinator of Language Services for an international faith-based organisation, I was hesitant towards the use of AI for translation purposes due to our specific terminology. However, we conducted a pilot project during summer 2025, implementing machine translation for an internal meeting”.
“The results confirmed our reservations. While the technology handled basic administrative content – agendas and procedural documents – adequately, it struggled significantly with ecumenical terminology, biblical quotations and doctrinal texts. In my opinion, machine translation currently remains insufficiently dependable, primarily due to the unpredictable nature of AI hallucinations. The continued involvement of experienced human translators proves essential for source verification and textual refinement” she says.
Yaroslav Makarevych, head of Action Global Communications Ukraine, a well-known PR agency, says: “We used AI tools for translating an annual report. However, no matter how many settings we applied, there were terms and phrases that changed. While they were grammatically correct and accurate, they were simply synonyms of terms used in previous years. AI could not identify where exact terms should be reused, which meant manual review was still necessary, so our language service partner had to check and correct everything almost from scratch.”
So, simple and straightforward tasks are generally manageable for AI under supervision, but as the complexity of the task increases with multiple steps, the risk also grows. Therefore, human oversight is crucial.
Client and market perspectives
Implementing AI technologies is neither cheap nor fast, especially when these systems cannot fully replace human work. For companies without a continuous, high-volume translation need, traditional human translation remains a more cost-effective option compared to investing heavily in AI integration.
Clients value AI for efficiency but continue to rely on LSPs for quality assurance and accountability. Strategic AI integration, combined with human expertise, meets client expectations while enabling faster and scalable operations.
Conclusion: AI as a core function in language services
In translation, within the last 10-12 years AI has evolved from an experimental add-on to a core operational function. Its strategic implementation demonstrates how firms can balance technological efficiency with human expertise. While AI handles routine and high-volume tasks, professional linguists provide oversight, cultural knowledge, and nuanced understanding that machines cannot replicate.
The recent AI hype has spotlighted these capabilities, but for industry insiders, this integration has been a long-standing reality. Moving forward, the challenge is refining AI workflows further—optimising efficiency while maintaining the irreplaceable value of human judgment. Strategic AI implementation ensures that technology and human expertise work together to meet the growing demands of global clients.
