Is machine translation penalised by Google?

What Google sanctions

Since March 2024, Google has used the term “scaled content abuse” for producing large numbers of pages with the primary purpose of manipulating rankings, without adding value for readers, no matter how they are produced. Among its examples: scraping content (feeds, search results…) and transforming it automatically, for instance by translating it, to turn it into lots of pages that offer readers next to nothing.

What is targeted is the purpose and the lack of value. Not the tool.

What Google says about automation

Google writes that appropriate use of automation is not against its guidelines. It asks you to take care of the accuracy, quality and relevance of the content, and to review what has been produced automatically, including titles, descriptions and image text.

And a translated page isn't a duplicate: Google only treats two versions as duplicates if the main content hasn't been translated.

In June 2025, Google even removed from its documentation the advice to block crawling of machine-translated pages, to bring it into line with its March 2024 rules.

Translating your own site isn't cheating

You're translating your pages for customers who speak another language. That's exactly what Google recommends: one version of each page per language, linked to the others.

The four precautions

1. Review the pages that matter: the home page, prices, booking or checkout.
2. Review the titles and descriptions shown in search results.
3. Translate the whole page, not just the menu and buttons.
4. Keep your proper names (your brand, your places) as they are.