Google has been using automatic artificial intelligence (AI) models to identify and eliminate the false reviews published on Google Maps, which has eliminated a total of 188 million deceptive reviews during the past year, while announced that it will begin to include warnings about “suspected reviews of being false” that have been identified and eliminated in companies.
Google Maps reviews help users to train an opinion about the place they are looking for. However, it is an option that can show false interactions by the users who write them, with invented reviews written by people who have never visited the establishment, or through positive reviews that are carried out under incentives such as discounts. Google does not allow the publication of any content that “does not reflect a genuine experience” in the place or establishment in question. In this sense, in order to identify false reviews and act quickly, Google has invested “considerably” in protection measures to eliminate them from their Google Maps service.
To do this, Google has been using automatic learning models, which are trained with broad data sets of known, both false and genuine data, facilitating the identification of patterns that indicate fraudulent activity. With what, in 2024, he eliminated a total of 188 million false and misleading reviews. Thus, as the technological explained in a statement in its transparency report and have collected media such as 9to5GOOGLE and Android Authority to identify the false content, the models used analyze various factors, such as the context of the account, with which they can detect possible connections with other suspicious accounts or paid review operations.
Likewise, the models can also identify patterns on the website of the place or establishment, as well as sudden inflows of reviews with five stars or situations with conflict of interests or incentives that may lead to biased reviews. Thus, since 9to5google they have detailed that, as they have known, Google has trained “a new model with the help of Gemini” that is capable of identifying potentially suspicious profile editions. As he explained, he only needs “a few examples” to identify what is suspicious and apply this knowledge to various business categories.
240 million eliminated reviews
In this way, with automatic learning models, Google blocked or eliminated a total of 240 million reviews during the past year because they violated their policies and, among them, 188 million were false reviews. Likewise, the company has also detailed that most of these reviews were eliminated before being seen by users. Following this line, Google has also detailed that it eliminated more than 12 million fake companies of Google Maps in 2024.
Among them, more than 10,000 profiles and false ads that belonged to a group of malicious actors tried to cheat people posing as real locksmiths. In addition to all this, Google has also reflected that its content reviewers team can evaluate reviews marked by the community to make decisions based on its policies. Thus, in the event that a company profile violates interaction policies with false users, Google recalled that it will be able to impose restrictions, in addition to eliminating false reviews. To this is added that, when a restriction is applied to a company profile, users will begin to see an alert on the page to inform that “reviews suspected of being false” have been identified and eliminated.
These warnings have begun to be implemented on Google Maps in the United States, the United Kingdom and India, although they will begin to extend globally during the next month.