For several years now, the language industry has lived under the reign of artificial intelligence (AI). No conference, trade report, or investor briefing seems complete without invoking its transformative power. Language industry analysts such as Nimdzi Insights and Slator dedicate entire issues to it; entrepreneurs promise AI-first solutions; and academics reinvent their research agendas around it. Whether by conviction or by necessity, everyone has become an AI expert, commentator, or even influencer.
This situation is certainly not unique to our field, but for language professions it runs particularly deep. The current AI wave is, after all, a language revolution. Large language models (LLMs) have placed linguistic competence at the core of technological progress. For translators and interpreters — who were once confident that machines would never master their art — the past two years have been a sobering revelation: Fluent machine translation (MT), constantly improving automated interpreting, and real-time dubbing and subtitling are all now part of everyday reality. Billions in investment have produced results once thought impossible. Businesses have quickly followed suit, seeking to leverage AI for a myriad of applications — sometimes with limited success, other times with impressive market penetration.
The language industry still seems dazzled by the momentum of AI, and understandably so, given the breathtaking pace of new breakthroughs. Impressive announcements keep coming, such as Meta’s ASR Omniscient and Translated’s Lara 200, which push speech recognition and multilingual LLMs to unprecedented coverage across low-resource languages — the true Achilles’ heel of the AI language revolution. So it is not surprising that hardly anyone in the field is asking, at least publicly, “What will happen if the AI bubble bursts?”
Lessons From the Last Bubble
The signs of a bubble are unmistakable, as many commentators are increasingly pointing at vast investments chasing limited (short-term) returns, speculative excitement masking the absence of clear business models, valuations untethered from revenue, and even financial tactics to artificially boost earnings. This is not the first time we’ve seen such exuberance. The dot-com era followed a clear trajectory: first, a frenzy of innovation and capital, then a spectacular crash, and, eventually, the steady, practical rebuilding that gave us the modern internet. With AI, we are now in the first phase, probably heading to the second.
While many might expect the bursting of the AI bubble to usher in another so-called “AI winter,” the reality could turn out quite different. What is important to remember is that the internet that rose from the wreckage of the dot-com crash became one of the great success stories of the early 21st century. For more than two decades, it reshaped commerce, media, communication, and business. Almost everything we rely on today depends, in one way or another, on the internet. We buy our vacations, refrigerators, and clothes online. Across much of the Western world, we depend on it for countless bureaucratic interactions with the state. We receive part of our education and information through podcasts, online courses, and scientific publications. Some even find love online.
The internet has, in practice, become what philosopher Luciano Floridi called the Onlife: our second digital life, intertwined with the material and analog one we have always known. The Onlife we are all part of happened after the dot-com crash. In some way, the bursting of the first technological bubble has made possible this digital revolution. Interestingly, the new reality that emerged was not much different from what the very dot-com companies that had once caused the crash had imagined.
The Language Industry’s Moment of Reckoning
If the AI bubble were to burst, the repercussions for the global economy could be far-reaching, and certainly not positive. The same holds true for the language industry, which has become deeply entangled in the AI narrative. When the bubble bursts, venture capital will dry up, projects will stall, and the generous ecosystem of “AI innovation funding” may shrink overnight. Startups will disappear, and mid-sized providers will struggle to compete with the few tech giants that manage to endure. We have seen this before.
Contrary to some overly optimistic predictions, and despite the temptation to welcome a market correction, such a collapse would not necessarily benefit the traditional language professions. Freelancers — the backbone of classic translation and interpreting — as well as academia, could face a second setback, compounding the challenges already introduced by automation and the rapid adoption of AI-driven tools.
And yet, it would be a mistake to view this as the end of opportunities in the language space, both for individual professionals as for companies, just as the dot-com crash was not the end of the internet or the industry it created. In fact, it marked the beginning of a new era where sustainable success was finally possible.
The Seeds Beneath the Ashes
As with the dot-com crash, what survives the collapse will matter more than what disappears. The technology is real. Its potential is enormous. Even without another paradigm shift, the applications of AI in language are only partially explored, and so are the business opportunities. More crucially, the infrastructure — data centers, foundational models, and the cultural familiarity with AI itself — will remain.
After the speculative froth evaporates, the industry will have a chance to rebuild on steadier ground. Companies that endure will be those that learn to integrate AI meaningfully rather than worship it.
Let’s be clear: Many will not survive; consolidation is likely; power will concentrate in the hands of a few large technology providers; and smaller players will depend on their platforms, data, and pricing. If the industry is not careful, the very diversity that makes human language vibrant could narrow into a few standardized pipelines controlled by corporate gatekeepers.
Space for a Fragile Optimism
It is difficult to make predictions, but we know that the bursting of the AI bubble will not mark the end of AI progress. Maybe it will mark the end of illusion. And consequently, the years ahead will be less euphoric, more pragmatic. That is not necessarily bad. Every technological revolution, once stripped of its hype, needs to find its social and economic equilibrium. And this is the time where market-fit and sustainable growth will take the place of bold initiatives and skyrocketing capital evaluation.This is the time for the makers.
For the language industry, the task is to prepare now for the bursting of the bubble — to understand how to thrive in an AI-first world without surrendering to it entirely. That means diversifying services, investing in human expertise where it still counts, and building new forms of collaboration between professionals, technology, and clients. And of course, build products that people really want to use.
We may not emerge unscathed, but we can emerge wiser. The dot-com collapse cleared the path for a more mature internet. The AI correction — whenever it comes — could do the same for our industry. The question is not whether it will happen, but whether we will be ready to build something lasting when it does.
