Google Takes Action Against Nonconsensual Deepfake Imagery
The rise of nonconsensual deepfake imagery on the internet, particularly targeting female celebrities, has posed a significant challenge for search engines. Even if users are not actively seeking explicit content, searching for certain names can lead to an alarming number of links to fake explicit photos and videos of those individuals. In response to this issue, Google has announced an update to its ranking systems aimed at addressing this problem.
In a blog post, Google product manager Emma Higham explained that the ranking updates are designed to reduce the visibility of explicit fake content in search results. When users search for terms associated with nonconsensual deepfakes of specific individuals, the ranking system will strive to provide “high-quality, non-explicit content,” such as news articles, whenever possible.
The aim of these changes is to ensure that people can learn about the impact of deepfakes on society without being exposed to actual nonconsensual fake images. According to Higham, these ranking updates have already resulted in a 70 percent decrease in the visibility of explicit image results on deepfake searches.
However, distinguishing between real and consensual content, such as an actor’s nude scenes, and deepfakes generated without consent can be challenging. To address this, Google will now consider whether a site’s pages have been removed from its search results due to policy violations when determining rankings. Sites with a high number of removals for fake explicit imagery will be demoted in search rankings.
Furthermore, Google is also updating its systems for handling requests to remove nonconsensual deepfakes from its search results. The goal is to streamline the process for victims to have their deepfakes removed. Once a victim successfully removes deepfakes of themselves from Google Search, the company’s systems will work to filter related results on similar searches and remove duplicates of the imagery.
Higham acknowledged that there is still more work to be done in combating nonconsensual deepfakes and that Google will continue to develop new solutions to assist those affected. This announcement from Google comes in response to the call made by the White House for tech companies to take action against the proliferation of explicit deepfake imagery.
In conclusion, Google’s efforts to update its ranking systems and combat nonconsensual deepfake imagery are a step toward creating a safer online environment. By prioritizing high-quality, non-explicit content in search results and demoting sites with a history of policy violations, Google aims to reduce the visibility of explicit deepfake content. Additionally, the improvements to the removal request process will make it easier for victims to have their deepfakes removed and related content filtered. While there is still more work to be done, Google’s commitment to developing new solutions demonstrates its determination to protect individuals affected by nonconsensual deepfakes.
