Google has introduced new measures to combat the spread of explicit deepfakes on its Search platform, aiming to make it more difficult for users to find manipulated images. As part of its ongoing efforts to tackle the issue, the company is enhancing its removal process for non-consensual fake imagery that features individuals.
While users have previously been able to request the removal of such images under Google’s policies, the company is now taking a more proactive approach. Whenever a removal request is granted, Google will also filter out similar explicit results from its search results, as well as scan for and remove duplicates of the offending image. This update is designed to alleviate concerns for victims who may worry about the same image resurfacing on other websites.
In addition, Google has updated its ranking algorithms to prioritize “high-quality, non-explicit content” when users search for explicit deepfakes featuring a specific individual. This means that if a user searches for explicit content related to a person, the results will instead surface news articles or other relevant content. Google also plans to educate users about the impact of deepfakes on society by displaying results that discuss the issue.
However, Google acknowledges that it still faces challenges in distinguishing between legitimate and fake explicit images. To address this, the company is demoting websites that have received a high volume of removal requests for manipulated images in Search. This approach has proven effective in the past for other types of harmful content, and Google believes it will help to reduce the visibility of low-quality sites that host deepfakes.
