Translate A Book With AI
Translate a Book
Toggle sidebar
Translating With AI Does Not Trigger AI Detectors
I ran 500 samples from 100 real book translations through Pangram. 495 scored as human written.
If a human writes text then translates it using an AI, will this be flagged as "AI content" by AI detection tools? This is question that matters a lot to me, since translateabook.com's goal is to produce high quality AI book translation that authors can actually use to publish in another language with minimum extra work. Triggering an AI-detector would be a signal that the AI translation process is smoothing over the author's voice, making it less publishable.
To put it to the test, I analyzed 500 random samples from 100 real-world book translations with Pangram.com, a currently quite popular AI detection tool.
The result, out of 500 samples:
2 were flagged as "mostly human, AI assisted"
1 was flagged as "AI assisted"
2 were flagged as "AI detected" (with a low AI score)
495 were marked as "Human written"
The short answer seems to be: No, AI translation doesn't trigger AI detectors - a encouraging sign for authors that AI translation can be a strong basis for publishing to new markets. Read on if you want more details on this experiment.
What Was Tested
I used translations coming from translateabook.com Standard Mode, which is the simplest translation mode and closest to a simple (big) prompt asking the AI to translate the text, with less input from the agent harness. Models used depended on the language pair, but were mostly Gemini Flash 3.1, DeepSeek v3 and ChatGPT 5.2.
I used samples ranging from 500 to 1000 words (or up to 2000 characters for languages with no spaces between words, like Chinese or Japanese). They were coming from documents in a variety of genres, including fiction and non-fiction, and a number of different languages.
Pangram displayed version is 3.3.2.
Because Pangram is somewhat expensive I ran only of few tests, where a number of variables (languages, genres, AI models...) are confounding each other and I would consider the results indicative only. However the results are so strongly in favor of "human written" that it does seem indicative that Pangram generally doesn't consider AI translations to be AI-written content.
The Results
Pangram gives a "Percent AI" score (like 4.8%) and a label (like "Human written"). I had a hard time finding an exact mapping between the two since I saw a few "Human written" with a higher AI percent score than some "Mostly human, AI assisted" labels, but it seems like roughly up to 20% is "Human written", 20-30% is "Mostly human, AI assisted" or "AI assisted" and above 30% is "AI detected".
As a comparison, if I enter the output of a chat session with claude or chatGPT into Pangram I get a score of 100% AI.
Here are the results per language :
Target language<br>Passages<br>Scored<br>Mean AI<br>Median AI
Dutch<br>115<br>115<br>1.4%<br>1.1%
German<br>73<br>73<br>0.8%<br>0.5%
Polish<br>67<br>67<br>1.5%<br>1.3%
English (US)<br>66<br>66<br>10.3%<br>9.2%
French (France)<br>25<br>25<br>1.3%<br>0.5%
Ukrainian<br>25<br>25<br>2.7%<br>0.9%
Bulgarian<br>21<br>21<br>0.6%<br>0.5%
Spanish (Latin America)<br>15<br>15<br>0.5%<br>0.4%
Czech<br>15<br>15<br>1.2%<br>0.7%
English (UK)<br>10<br>10<br>0.8%<br>0.6%
Swedish<br>10<br>10<br>2.5%<br>1.1%
Russian<br>10<br>10<br>9.0%<br>5.2%
Finnish<br>4.4%<br>4.2%
Portuguese (Brazil)<br>0.9%<br>0.3%
Vietnamese<br>2.4%<br>1.3%
Turkish<br>0.6%<br>0.3%
Greek<br>7.1%<br>0.5%
Uzbek<br>0.4%<br>0.4%
Hebrew<br>0.6%<br>0.5%
Italian<br>0.4%<br>0.4%
Spanish (Spain)<br>0.3%<br>0.3%
Romanian<br>0.9%<br>1.0%
Chinese (Simplified)<br>2.7%<br>2.7%
All the "Percent AI" scores are really low - US English is higher, maybe Pangram is more sensitive there (though not in UK English?).
Pangram says they currently support the following languages: English, Arabic, Chinese, Czech, Dutch, French, German, Greek, Hindi, Hungarian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Turkish, Ukrainian, Urdu, and Vietnamese.
So actually Bulgarian, Finnish, Uzbek and Hebrew where in the sample are not officially supported by Pangram. This represents 37 samples - checking my results, all of them were scored as "Human written".
In case you're curious, "AI detected" were a Greek and a Russian samples, "Mostly human, AI assisted" were a Russian and a Ukrainian samples, and "AI assisted" was a US English (US) sample.
Here are the results per genre :
Book type<br>Passages<br>Scored<br>Mean AI<br>Median AI
Novel<br>342<br>342<br>2.3%<br>0.9%
Novella<br>28<br>28<br>2.6%<br>0.7%
Academic<br>25<br>25<br>3.0%<br>1.5%
Memoir<br>23<br>23<br>2.1%<br>0.8%
Spiritual<br>20<br>20<br>2.2%<br>1.5%
Short Stories<br>15<br>15<br>12.7%<br>15.8%
Science<br>11<br>11<br>7.2%<br>7.2%
Biography/Autobiography<br>10<br>10<br>0.7%<br>0.3%
Self-Help<br>10<br>10<br>2.7%<br>1.1%
Textbook<br>0.9%<br>0.3%
Children<br>0.7%<br>0.6%
Manual<br>0.4%<br>0.4%
Health<br>3.6%<br>3.6%
All still very low. "Short Stories" was the highest, and when I check this was mostly English (US). The language and genre variables are confounding each other so we can't draw too strong a conclusion, but it seems like the AI score stays quite low across genres. Interesting, given that...