Freelance translators have lost up to 80% of their income since ChatGPT launched, and the profession is collapsing faster than any other creative field

Translation was the first creative profession to be hollowed out by AI. More than a third of translators reported losing work due to generative AI by 2025, with 43% reporting income drops. Some translators have seen earnings fall by 60-80% from their peak. The International Monetary Fund cut its translator and interpreter staff from 200 to 50. One translator told CNN he lost 70% of his income when EU translation work dried up. The American Translators Association reports that translators are leaving the profession in significant numbers. The reason translation collapsed first is that translation has a uniquely verifiable output — a correct translation can be checked against the source, making it easy for clients to validate AI output and decide human translators are unnecessary. But what clients do not realize is that AI translation still fails catastrophically on context, cultural nuance, legal precision, and literary voice. The problem is that these failures are invisible to monolingual clients who cannot read the source language. A marketing team that uses AI to translate campaign copy into Japanese may never know that the translation is tonally wrong, culturally offensive, or grammatically awkward — they just see that it is cheap and fast. By the time the damage shows up in poor market reception or customer complaints, the translators who could have prevented it have already left the profession. This problem persists because the remaining work for human translators is increasingly degrading. Instead of translating from scratch, translators are now hired as 'post-editors' — cleaning up AI output at rates 50-70% lower than original translation rates. This means translators are doing the cognitively hardest part of the work (catching subtle errors that AI misses) while being paid less than ever. The economic incentive to enter the profession has evaporated: translation programs at universities are seeing declining enrollment, and the pipeline of skilled translators for less-common language pairs is drying up. Oxford researchers estimate that roughly 28,000 additional translator jobs would have existed in the US alone without machine translation's impact.