For twenty years, my work has depended on precision — legal filings, tribunal transcripts, humanitarian reports where a mistranslated term isn’t just an error, it’s a liability. I’ve never been an easy convert to shortcuts.
But Trados Copilot has genuinely changed how I work — not by replacing judgment, but by clearing the noise around it.
Terminology suggestions surface inline instead of sending me hunting through a glossary. Auto-suggest handles repetitive segments before I’ve finished typing them. QA checks catch small inconsistencies before they reach a reviewer. What that frees up is attention for what actually matters: nuance, register, the cultural weight a phrase carries in Kinyarwanda that it doesn’t in French or English.
AI in translation isn’t a shortcut around expertise — it’s a multiplier for it. A junior translator using Copilot still needs the judgment to know when a suggestion is wrong. That judgment isn’t something the tool teaches.
The translators who’ll thrive next aren’t resisting these tools, or blindly trusting them. They’re the ones who know exactly where the machine’s confidence ends and their own has to begin.
Curious how other language professionals are using AI in their CAT workflows — has it changed anything real for you, or is it mostly noise?