AI speaks every language. So why learn French anymore?
Artificial intelligence can now translate emails, meetings and conversations in seconds. Students are using chatbots to write in Spanish and French, while businesses increasingly rely on AI tools for multilingual communication. As instant translation becomes common, universities are confronting a difficult question: if AI already speaks every language, why should students still spend years learning one?
A report by QS Insights Magazine says the debate is reshaping language education globally as universities face falling enrolments, budget pressure and rapid advances in AI translation tools.
Language learning was already declining
The decline in language learning started well before ChatGPT.
The latest data from the Modern Language Association (MLA), released in late 2023, showed enrolments in languages other than English at US colleges and universities fell 16.6 percent between 2016 and 2021, the sharpest decline recorded in the organisation’s survey history. The MLA said enrolments have fallen 29.3 percent since their peak in 2009 and returned to levels last seen in 1998.
French enrolments fell 23.1 percent between 2016 and 2021, while German declined 33.6 percent, according to MLA data. In contrast, Korean enrolments rose 38.3 percent, reflecting growing student interest in globally influential Asian languages.
The decline was particularly sharp at two-year institutions, where language enrolments dropped 24.2 percent during the same period.
Experts interviewed by QS Insights Magazine said AI is accelerating an existing crisis rather than creating a new one.
Jen William, Professor of German at Purdue University and Co-Chair of the Modern Language Association’s Task Force on World Languages and Generative AI, said fewer students are now reaching university with strong language foundations because of reduced school-level language teaching.
Joseph Lo Bianco, Professor Emeritus in Language and Literacy at the University of Melbourne, said funding priorities and the growing focus on STEM and business programmes remain bigger drivers of decline than AI itself.
AI is making translation easier
AI tools are rapidly changing how students and employers view language learning.
Machine translation systems now offer fast and low-cost communication across multiple languages. Businesses can translate documents instantly, while travellers can hold real-time conversations using AI-powered apps. For many users, the technology is becoming sufficient for everyday communication.
Philipp Koehn, Professor at Johns Hopkins University whose work focuses on machine translation, told QS, AI tools make communication easier but may also encourage people to rely on technology instead of learning languages deeply.
The rise of generative AI has also changed student behaviour. Universities report growing use of chatbots for writing assistance, translation and grammar correction in foreign-language assignments.
At the same time, AI is creating a paradox for language education. While translation tools reduce the practical need for basic language skills, they also expose more people to foreign languages and cultures through global media, entertainment and online communication.
What AI still cannot do
Experts argue that translation is only one part of language learning.
Ana Niño, Senior Lecturer in Spanish at the University of Manchester, told QS Insights Magazine that language education develops intercultural understanding, communication skills and critical thinking in ways translation software cannot replicate.
Researchers say AI systems still struggle with humour, emotion, slang and cultural nuance. These limitations become more serious in diplomacy, healthcare, law and international business, where mistranslation can alter meaning and damage trust.
William also warned that AI-generated text can reproduce cultural stereotypes and bias, making human understanding increasingly important.
Experts say language graduates are often better equipped to identify social context, political sensitivity and hidden meaning in communication — areas where AI systems continue to face limitations.
UNESCO has also warned that many of the world’s languages remain underrepresented online, creating uneven performance in AI translation systems. As a result, AI tools work far better in dominant languages such as English, French and Spanish than in many regional and Indigenous languages.
This raises concerns that growing dependence on AI translation could create a more English-centric global system in which smaller languages face increasing digital exclusion.
Universities are adapting, not abandoning languages
Despite the wider decline, some universities are changing how language education is delivered rather than abandoning it altogether.
The MLA found that 38.3 percent of language programmes either maintained or increased enrolments between 2016 and 2021, particularly programmes linked to business, technology, healthcare and international relations.
Combined degrees that integrate language learning with professional fields are becoming more common in the UK, US and Australia. Universities are increasingly promoting language study as a workplace and global communication skill rather than only an academic subject.
Experts say the future of language education may depend less on producing translators and more on preparing students to work across cultures in an AI-driven world.
The debate, they argue, is no longer about whether AI can translate French correctly. It is whether universities still see value in teaching students to understand the people speaking it.
A report by QS Insights Magazine says the debate is reshaping language education globally as universities face falling enrolments, budget pressure and rapid advances in AI translation tools.
Language learning was already declining
The decline in language learning started well before ChatGPT.
The latest data from the Modern Language Association (MLA), released in late 2023, showed enrolments in languages other than English at US colleges and universities fell 16.6 percent between 2016 and 2021, the sharpest decline recorded in the organisation’s survey history. The MLA said enrolments have fallen 29.3 percent since their peak in 2009 and returned to levels last seen in 1998.
French enrolments fell 23.1 percent between 2016 and 2021, while German declined 33.6 percent, according to MLA data. In contrast, Korean enrolments rose 38.3 percent, reflecting growing student interest in globally influential Asian languages.
The decline was particularly sharp at two-year institutions, where language enrolments dropped 24.2 percent during the same period.
Experts interviewed by QS Insights Magazine said AI is accelerating an existing crisis rather than creating a new one.
Jen William, Professor of German at Purdue University and Co-Chair of the Modern Language Association’s Task Force on World Languages and Generative AI, said fewer students are now reaching university with strong language foundations because of reduced school-level language teaching.
Joseph Lo Bianco, Professor Emeritus in Language and Literacy at the University of Melbourne, said funding priorities and the growing focus on STEM and business programmes remain bigger drivers of decline than AI itself.
AI is making translation easier
AI tools are rapidly changing how students and employers view language learning.
Machine translation systems now offer fast and low-cost communication across multiple languages. Businesses can translate documents instantly, while travellers can hold real-time conversations using AI-powered apps. For many users, the technology is becoming sufficient for everyday communication.
Philipp Koehn, Professor at Johns Hopkins University whose work focuses on machine translation, told QS, AI tools make communication easier but may also encourage people to rely on technology instead of learning languages deeply.
The rise of generative AI has also changed student behaviour. Universities report growing use of chatbots for writing assistance, translation and grammar correction in foreign-language assignments.
At the same time, AI is creating a paradox for language education. While translation tools reduce the practical need for basic language skills, they also expose more people to foreign languages and cultures through global media, entertainment and online communication.
What AI still cannot do
Experts argue that translation is only one part of language learning.
Ana Niño, Senior Lecturer in Spanish at the University of Manchester, told QS Insights Magazine that language education develops intercultural understanding, communication skills and critical thinking in ways translation software cannot replicate.
Researchers say AI systems still struggle with humour, emotion, slang and cultural nuance. These limitations become more serious in diplomacy, healthcare, law and international business, where mistranslation can alter meaning and damage trust.
William also warned that AI-generated text can reproduce cultural stereotypes and bias, making human understanding increasingly important.
Experts say language graduates are often better equipped to identify social context, political sensitivity and hidden meaning in communication — areas where AI systems continue to face limitations.
UNESCO has also warned that many of the world’s languages remain underrepresented online, creating uneven performance in AI translation systems. As a result, AI tools work far better in dominant languages such as English, French and Spanish than in many regional and Indigenous languages.
This raises concerns that growing dependence on AI translation could create a more English-centric global system in which smaller languages face increasing digital exclusion.
Universities are adapting, not abandoning languages
Despite the wider decline, some universities are changing how language education is delivered rather than abandoning it altogether.
The MLA found that 38.3 percent of language programmes either maintained or increased enrolments between 2016 and 2021, particularly programmes linked to business, technology, healthcare and international relations.
Combined degrees that integrate language learning with professional fields are becoming more common in the UK, US and Australia. Universities are increasingly promoting language study as a workplace and global communication skill rather than only an academic subject.
Experts say the future of language education may depend less on producing translators and more on preparing students to work across cultures in an AI-driven world.
The debate, they argue, is no longer about whether AI can translate French correctly. It is whether universities still see value in teaching students to understand the people speaking it.
Next Story