With the rise of AI and machine translation, many agencies and freelancers have started offering MTPE (Machine Translation Post-Editing) at lower prices and… with lower quality too. 🤖
I stay true to my values and never use AI to translate, transcreate, or write. Here’s why:
- Quality suffers. AI doesn’t understand context, culture, or humor. Even after “post-editing,” the result will never sound as natural and persuasive as a 100% human translation. Over time, AI-generated content will fill the web with the same repetitive language and tone, a true marketing nightmare.
- It’s a copyright and privacy risk. AI tools reuse existing content written by professionals without consent. That’s not only unfair: it’s often illegal. Plus, when you input your confidential data into AI systems, you lose control over where it goes.
- It’s unfair to professionals. Editing machine-translated text takes longer and pays less than translating from scratch. It’s demotivating, inefficient, and unsustainable. I’ve tried to edit some machine-translated texts. It took me twice as long than translating them from scratch. Same for copywriting.
👉 Moral of the story: Nobody wins with AI.

The risks of AI + human editing
I will never work with AI. Here’s my story.
For over 6 months, I’ve been reviewing a large batch of AI-generated translations that had already gone through “human revision”.
What surprised me was not the AI output itself (awful, of course, but I was expecting that). It was how many obvious errors survived the review stage.
There were 4 reviewers, all of them native, professional translators and copywriters. I read an article corrected by one of them. Weird. It looked like they didn’t change anything, it still sounded AI-generated and there were many mistakes (grammar, punctuation, style). So I decided to read more of them. I read 5, 10, 15… All of them looked like no human reviewed them.
With the client we agreed I would review all of them. Hundreds of articles. And ALL of them were the same.
Here are some mistakes the reviewers didn’t spot:
❌ Calques from English
❌ Terminology errors
❌ Literal translations of idioms (that made no sense in the target language)
❌ Proper nouns and names wrongly rendered
❌ Localization failures (currencies, date and time, cultural references)
❌ Tone-of-voice inconsistency
❌ Gendered language (masculine instead of gender-neutral alternatives)
❌ Overly dramatic or artificial language (typical of AI)
And I could go on…
So, what happened? Why did four professional translators/copywriters did such a bad job?
This is exactly where AI can become dangerous in translation workflows: reviewers stop reading critically and end up reading too quickly, too trustingly, or without the level of attention they would normally apply to a human translation from scratch.
When editing, they are no longer using their own creativity and logics. They mimic the AI one. And the quality of the final text is inevitably bad.
What does this mean for the client? That less people will buy their products and services, because they don’t trust them.
So, what’s the solution? 🧚
Always choose Human translation + human revision process. This workflow is the one that produces the strongest quality.
That’s what you get when you work with me.
I collaborate with trusted colleagues so we can provide a translation + review workflow. 100% human, 100% top-notch quality.
If you need an Italian translator or reviewer (or both), you know what to do: get in touch with me at morethantranslation.mtt@gmail.com