Google announced it is taking action against thousands of what it termed bad apps on its Play Store and removing millions of fake reviews, through an artificial intelligence (AI)-powered anti-spam system.

In a blog post, the company said its Trust and Safety team deployed a system combining human intelligence and machine learning earlier in the year to detect and enforce policy violations in the ratings and reviews of apps.

The company took the measures after finding ratings and reviews of its apps could have been violating guidelines, thus undermining user trust.

In particular, Google said its Play Store was being subject to bad content – reviews that were hateful or off topic – along with fake ratings to manipulate an app’s average. It was also plagued with incentivisation issues, whereby people gave ratings and reviews in exchange for money or valuable terms.

As part of the crackdown, a team of engineers and analysts closely monitored and studied suspicious activities in Play Store’s ratings and reviews, thus improving the model’s precision.

In the past week, the company said millions of reviews and ratings had been detected and removed from the Play Store, while thousands of bad apps had been identified due to suspicious reviews and rating activities on them.

As part of the effort, the company said developers should avoid buying fake or incentivised ratings and not run campaigns in apps.

Users were also urged not to accept money or goods in exchange for reviews and ratings, or to use profanity to criticise an app.