The rise of explicit deepfakes has posed a significant challenge, affecting the privacy of individuals worldwide. Deepfakes are AI-generated images or videos depicting someone in compromising situations without their consent. Such content has become increasingly prevalent. These explicit fakes often appear in search results, making it difficult for victims to control their online presence.
In response to this growing concern, Google is rolling out new safety features designed to tackle explicit deepfakes more effectively. The new measures simplify the process of removing explicit non-consensual fake content from Search.
When users successfully request the removal of such content, Google's systems will now also filter out similar explicit results and remove duplicate images. This approach aims to prevent the reappearance of unwanted content.
'These protections have already proven to be successful in addressing other types of non-consensual imagery, and we’ve now built the same capabilities for fake explicit images as well,' said Emma Higham, Google product manager.
The goal is to offer peace of mind to individuals concerned about future occurrences of similar content. Google is also adjusting its Search rankings to better handle high-risk queries. Searches specifically looking for deepfake images of real people will now show high-quality, non-explicit content, such as relevant news stories.
Additionally, websites with numerous removals for fake explicit imagery will see their rankings lowered. Previous updates have already reduced exposure to explicit image results on deepfake-related queries by over 70 per cent this year.
Google is also developing methods to distinguish between real explicit content, such as consensual nude scenes, and explicit fake content, ensuring that legitimate images remain accessible while demoting deepfakes. These updates are part of Google’s ongoing efforts to manage explicit and dangerous content online.
In May, Google banned advertisers from promoting deepfake porn services. In 2022, it expanded the types of 'doxxing' information that can be removed from Search. Additionally, in August 2023, Google began blurring sexually explicit imagery by default. By continuously enhancing its online safety features, Google aims to protect users from harmful and explicit content, ensuring a safer and more secure online experience.