Will AI replace translators? What the evidence shows
Machine translation is good enough that companies use it for routine content and freelance translators report lost work and income. For high-stakes text and live interpreting, studies still find errors that need professional review, and BLS projects small growth rather than decline.
The job today
Task by task: what AI can actually do
In real use means deployed at scale; shown in tests or pilots means demonstrated but not yet routine; not yet means we found no credible evidence it does this reliably.
| Task | Status | Evidence |
|---|---|---|
| Translating routine product and app content | In real use | Duolingo cut about 10% of its contractors in Jan 2024 as it used more AI to create content, saying humans still review AI output. (Semafor, Jan 2024) |
| Marketing and corporate translation | In real use | Klarna reported cutting spending on external translation and other agencies, part of a $4 million run-rate saving from generative AI (May 2024). (Marketing Dive, May 2024) |
| Translating medical instructions | Shown in tests or pilots | UCSF researchers found GPT-4 translated ER discharge instructions with 97% sentence accuracy in Spanish but 89% in Russian, where 56% of instruction sets had an error; they advise professional review for high-stakes content. (BMJ Quality & Safety, 2026) |
| Live interpreting in courts, hospitals and negotiations | Not yet | We found no rigorous study showing AI interpreting is reliable in these settings; BLS says many interpreter and translator jobs cannot yet be entirely automated. (BLS OOH, 2025-35 projections) |
Jobs evidence so far
Labelled by who says AI was the reason: the company itself, reporters, or a study.
- Oxford study: each 1-point rise in machine-translation use is linked to ~0.7 points slower translator employment growth, roughly 28,000 positions not created in 2010-2023 (study) (CEPR VoxEU, Mar 2025)
- Society of Authors survey of 787 members: over a third of translators had lost work to generative AI and about 40% said it reduced their income (study) (The Bookseller, Apr 2024)
- Duolingo cuts about 10% of contractors as it uses more AI; it said calling AI the sole reason would be an 'oversimplification' (company says AI, in part) (Semafor, Jan 2024)
- Microsoft's study of 200,000 Copilot conversations puts interpreters and translators at the top of its AI-applicability list, with 98% of their activities overlapping AI tasks (study) (Sinclair/Newschannel 9, Aug 2025)
What to do
- Move to high-stakes specialismsLegal, medical, certified and regulated translation still requires professional review, and errors carry liability.
- Offer post-editing and quality reviewMuch routine work now starts as machine output; translators who can review and certify it quickly stay in the workflow.
- Build interpreting and cultural skillsLive, in-person interpreting and transcreation for marketing rely on context machines handle poorly.
- Track rates, not just volumeFreelancers report income falling before work disappears; watch per-word rates and client mix closely.
Questions people ask
Will AI replace translators?
It is already reducing demand for some routine translation, and research links machine translation to slower translator job growth. BLS still projects 2% growth to 2035 and says many jobs cannot be fully automated.
Is AI translation accurate enough for medical use?
For common languages like Spanish it is often accurate at the sentence level, but a 2026 study found errors in over half of Russian instruction sets and recommends professional review for high-stakes content.
Which translation tasks are safest?
Live interpreting in legal and medical settings, certified translation, and creative or cultural adaptation.
Have translators actually lost work to AI?
Over a third of translators in a 2024 UK Society of Authors survey said so, and about 40% reported lower income.
How we grade
Each task is graded 'deployed' only when a named organisation reports AI doing it in real work at scale, 'demonstrated' when it works in tests, benchmarks or pilots, and 'not yet' when we found no credible evidence AI does it reliably. For job losses we separate what a company said in its own words ('company says AI') from what reporters or analysts attributed to AI ('reported as AI') and from research ('study'); employment and outlook figures come from the BLS 2025-35 projections.
Sources
- BLS Occupational Outlook Handbook: Interpreters and translators
- CEPR VoxEU: Lost in translation (Frey & Llanos-Paredes)
- BMJ Quality & Safety: machine translation of discharge instructions
- Semafor: Duolingo contractor cuts
- The Bookseller: SoA translator survey
- Marketing Dive: Klarna generative AI savings
- Microsoft: Working with AI (arXiv)
Spot an error or something out of date? Tell us; corrections are made and logged. Written with AI assistance from public sources; every fact checked and approved by Colin Balfe. General information, not legal, medical or financial advice.