SPECIAL FEATURE: AI IN T&I (Part 4)
With AI use increasing at an exponential rate in practically all fields, AUSIT National President Carl Gene Fordham surveys recent media articles on the topic to find out how this major technological shift is affecting translation and interpreting across industries.
… the $15.8 billion interpreting market is undergoing genuine disruption …
AI-powered translation is reshaping communication across industries at a speed and on a scale that was unimaginable just a few years ago.
In May this year, cloud-based AI translation and captioning platform Wordly announced that it had reached the milestone of over one billion minutes of AI-generated translation across 120 countries.1 This achievement signals that the $15.8 billion interpreting market is undergoing genuine disruption, with multilingual support now accessible to event organisers who could never have afforded human interpreters at scale.
Yet the promise of AI translation is most compellingly tested not in conference halls, but in settings where the stakes are far higher. In surgical wards at institutions like Brigham and Women’s Hospital (Boston, USA), for example, researchers have found that patients want AI and human interpreting to coexist rather than compete – leaning on technology for speed and convenience in routine exchanges, while turning to human interpreters when empathy and cultural nuance are indispensable to their care.2
As I write this article, a similar tension is set to play out on the world’s biggest sporting stage: the 2026 FIFA World Cup, in which 48 teams, three host nations and over 104 matches will create a linguistic challenge that the international football association’s ageing communications infrastructure is ill equipped to handle. From multilingual dressing rooms to error-prone press conference interpreting – the kind that once sent Chelsea’s transfer news into chaos over a confusion between ‘midfield’ and ‘defence’ – the tournament lays bare just how far the sport has fallen behind.3
Taken together, these three arenas tell a consistent story: AI translation is not a wholesale replacement for human interpreters, but it is fast becoming an indispensable layer in any serious language access strategy, expanding what is possible – while the most complex, emotionally charged and high-stakes conversations continue to remind us why the human element still matters.
AI is increasingly being applied to high-stakes T&I contexts, yet its adoption raises profound questions about accuracy, cultural understanding and human dignity. In asylum systems, for example, the Lowy Institute has found that AI’s pattern recognition is a poor substitute for human understanding, and as a result AI translation tools risk misrepresenting the nuanced personal testimonies of some of the world’s most vulnerable people, with potentially life-altering consequences for those who lack the means to challenge technological errors.4
In the realm of human interpreting more broadly, the rise of real-time AI voice interpreting threatens to render obsolete a centuries-old tradition that fostered not just communication but genuine cross-cultural curiosity and understanding – reducing language to mere functional code.5
Meanwhile, AI tools designed to translate American Sign Language, while positioned as accessibility solutions, have been criticised for missing critical linguistic features such as facial expression, raising serious doubts about whether they serve the Deaf community’s needs or simply create the appearance of inclusion.6
Yves Valerus, a Haitian Creole–English interpreter and single mother of three based in Brooklyn, New York, saw her income drop by nearly 20% after her employer introduced new scheduling software that fragmented her previously stable hours – forcing her to make difficult choices such as prioritising her internet bill over utilities and travelling further to find groceries on sale. In response, Valerus and colleagues are attempting to unionise with the Communications Workers of America, with an additional concern being their employer’s announcement that it is exploring use of AI to handle basic interpreting work.
The situation these employees find themselves in illustrates the broader trend of algorithmic scheduling – technology designed to minimise labour costs – which has spread across industries over the past decade, leaving hourly workers with erratic and unpredictable hours. Labour organisers are now drawing on that experience to warn that generative AI in the workplace poses similar risks, and that workers must have a voice in how new technologies are introduced.7
Carl Gene Fordham is a NAATI-certified translator (Chinese>English) and interpreter (Mandarin–English) based in Melbourne. He currently serves as AUSIT’s National President.
1. https://www.meetings-conventions-asia.com/News/Technology/Wordly-passes-1-billion-minutes-of-AI-live-translation
2. https://www.digitaljournal.com/article/new-study-examines-ai-medical-interpreter-use-for-surgical-care/
3. https://breakingthelines.com/@btl/2026-world-cup-48-teams-dozens-of-languages-one-tournament
4. https://www.lowyinstitute.org/the-interpreter/high-stakes-environment-ai-failing-asylum-seekers
5. https://www.theguardian.com/world/commentisfree/2026/may/09/ai-interpretation-diego-marani
6. https://www.nbcboston.com/news/local/ai-deaf-americans-asl-interpreter/3935273/
7. https://www.wuft.org/2026-05-03/how-algorithms-wreaked-havoc-with-these-workers-schedules-and-cut-their-pay?_amp=true
Advertisement:
