Chartered Institute
of Linguists

Are we in a (Language) AI Bubble?

 


By Dom Hebblethwaite, CIOL Head of Membership 

 

Our latest CIOL poll asked members a simple but provocative question: are we in a language AI bubble? The responses reveal a profession navigating rapid technological change and hype around AI with strongly held views about the value of human expertise. 

 

The numbers 


Especially for translators, hype around technology starting with Machine Translation is nothing new. As with most polls around AI and technology, there are a range of views. Nearly half of respondents (49%) believe the current hype around AI exceeds its real-world utility. Just over a quarter (28%) feel the buzz broadly matches AI's potential, while almost a quarter (23%) remain unsure, suggesting it's too early to tell. These figures point to a profession that is far from unified in its assessment of AI's transformative claims. 

 

Experience from the front lines 


For many members, scepticism about AI is rooted in direct professional experience. One translator with over thirty years in the industry offered a sobering assessment: 
 

"I have seen AI erode quality of translation, workflow, rates, etc. AI output for quality work requires much more work than it would take to translate from scratch. Post-editing AI output is tedious, counter-creative and deadening. It's akin to data entry and is being paid accordingly." 
 

This sentiment was echoed by a translator with six years' experience: 
 

"I've seen this excessive use of AI as an alternative to translators and linguists, when in fact it's extremely flawed and heavily reliant on human input. And all to save a few bucks and try to kick language experts to the curb if they don't accept pennies as payment for that input." 

 

The perception gap 


Several respondents highlighted what might be called a perception gap: a disconnect between how AI is marketed and how it actually performs in specialist contexts. As one member put it: 
 

"The people promoting AI as a solution for 'everything' don't know anything about the subjects they claim it can be used for, including language. Translation is seen as 'simply' replacing words, not as recreating a message in a different language against a different cultural background." 
 

One retired Associate member tested AI capabilities in a rather different domain (cryptic crosswords) and found the results wanting: 
 

"An example is: 'Doctor finds article in drain.' 5 letters. It went through a long-winded argument about doctors using stethoscopes and came up with 'Steth'. Artificial logic, yes, but intelligence? Zero. Sometimes it even got the number of letters wrong." 
 

Signs of a turning tide? 


Some members report signs that clients are beginning to recognise AI's limitations. A creative translator and games localisation specialist observed: 
 

"A lot of what's shaped my view is the various language companies I'm seeing that have tried AI for a period and had to go back to more human-centred processes, citing a significant loss of quality and revenue." 
 

Another member expressed hope that 2026 might bring a recalibration: 
 

"AI has had a direct, huge impact on my workload, now down to about 20% of what it was in 2024. I'm hoping 2026 will see clients come to their senses, realising that even AI translations need checking by human translators." 


Existential questions 


 Not all members see AI as overhyped. One argued the opposite: 
 

"AI's potential is larger than the buzz, not smaller or broadly equal. It ranks with humanity's most important discoveries and inventions: the wheel, electricity, writing, airplanes. But it will rise to the top in those rankings, and have the power to wipe out humanity which previous inventions did not have." 
 

Others took a more defiant stance. As one member put it bluntly:  
 

"No bloody machine can EVER replace and be as accurate as my brains." 


Looking forward 


Nearly half our respondents believe we are indeed in a bubble and that the promises of AI have outpaced its delivery, at least for now. 
 

Yet a significant minority do not believe AI is overhyped, and many acknowledge we simply don't yet know how this will play out. 
 

What's clear is that our members are not passive observers. They're adapting, questioning, and in many cases pushing back against narratives that undervalue human expertise.  
 

Whether the AI bubble bursts, slowly deflates, or proves to have been no bubble at all, CIOL will continue to advocate for professional standards and the irreplaceable value that skilled linguists bring to interpersonal and cross-cultural communication. 

 

 

Dom Hebblethwaite is Head of Membership for the Chartered Institute of Linguists. For more on Dom see his profile here.  

 

 

 

 

 

 

Views expressed on CIOL Voices are those of the writer and may not represent those of the wider membership or CIOL.