This is a burning question for today’s translation industry. Let’s put it to the jury — you, the readers.

Case for the Prosecution: AI

The argument goes something like this:

• AI tools like DeepL and Systran can translate vast amounts of content in seconds
• They speed up translation and offer multiple suggestions
• Particularly useful for translating technical/medical abbreviations
• Tools like ChatGPT and Consensus can instantly assist with terminology research

Looks like a slam dunk for AI, right?

Case for the Defence: The Human Translator

Here’s why the human translator still matters — and why fair pay must remain the norm:

• Education: Professional translators often hold BA Honours or Master’s degrees — an investment that should be appropriately compensated
• Most source documents are in PDF format and need to be converted and reformatted — a time-consuming task that AI can’t handle on its own
• Raw machine translation output requires detailed human revision:
  • Consistency of terminology
  • Technical accuracy
  • Correct date formats (UK vs US)
  • Repetitions, omissions, and mistranslations
  • Proper names often get wrongly translated
• Expectations have changed — clients now demand lightning-fast delivery, increasing pressure on translators

Real-Life Case Studies

Medical Reports
I was recently sent a batch of AI-translated medical reports on alcohol-induced cirrhosis of the liver. The PDFs had been poorly converted, resulting in garbled formatting. AI completely failed to grasp the technical content — the output was unintelligible. I had to reformat everything and essentially retranslate from scratch.

General Contracts
I’ve had more success with AI-generated translations of general agreements — but even here, careful review is essential. I’ve had to:

• Ensure consistency in terms like “contract” vs “agreement”
• Remove repeated content within paragraphs
• Correct translated proper names
• Double-check all figures

The Verdict?

While AI dramatically accelerates the initial phase of translation and saves time on research, it doesn’t eliminate the translator’s role — it shifts it.

Translators still:

• Reformat source files
• Thoroughly revise AI output
• Take responsibility for accuracy
• Work faster than ever before, often under unrealistic deadlines

And yet — this hidden labour is rarely recognised or fairly paid.

Over to You, the Jury…

So what do you think? Should translators be paid less in the age of AI — or should their expertise be valued more than ever?