Google strengthens search security measures to combat malicious deepfake content

Google is actively launching a series of innovative online security measures to greatly simplify and accelerate the process of removing malicious deepfake content from search engines, while preventing such inappropriate content from occupying prominent positions in search results.

When users successfully apply to remove unauthorized and explicitly false images related to their own image from Google search, Google's system will further expand the scope of protection, automatically screen and clean up any similar inappropriate results that may appear in all related searches, and delete any duplicate violations images, providing users with more comprehensive privacy protection.

Emma Hayem, a product manager at Google, emphasized in the announcement: "We have successfully applied these protective measures to deal with other types of unauthorized image issues, and now we are applying the same technology to combat explicit false images. The core of these efforts is to give people more security, especially when they are concerned about more inappropriate content about themselves on the future Internet."

In addition, Google has also adjusted its search ranking algorithm to more effectively deal with high-risk queries that may breed a large number of malicious deepfake content. For example, for queries deliberately searching for deepfake images of public figures (such as the previously circulated AI-generated sexual images of Taylor Swift), Google will prioritize displaying "high-quality, non-explicit content," such as relevant news reports, to reduce potential harm to users. For websites that frequently publish such violations, Google will take measures to lower their search rankings.

According to Google, this year's series of updates have significantly reduced the appearance rate of explicit images in search results specifically searching for deepfake content by more than 70%. At the same time, the company is actively exploring technical means to distinguish legitimate explicit content (such as nude scenes filmed with the consent of actors) from malicious false images, aiming to purify the online environment while ensuring the normal display of legitimate content.

This series of updates is part of Google's extensive actions to address dangers and inappropriate content in the online space. In May of this year, Google banned advertisers from promoting any deepfake pornography services. As early as 2022, Google expanded the scope of "doxxing" information that can be removed from search, and starting from August 2023, it will blur explicit sexual images by default to further protect user privacy and dignity.