Search Engine Update: Google Works to Suppress Explicit Deepfakes

Google is set to roll out a series of new online safety features aimed at bolstering user protection against explicit deepfakes, particularly in its search results. This initiative seeks to streamline the process for individuals wishing to remove explicit, nonconsensual deepfake content that depicts them, making it more effective and comprehensive.

When a user successfully requests the removal of explicit imagery that is a deepfake, Google's systems will now automatically filter out all explicit results related to similar searches about that user. This means that if a specific deepfake image is flagged and removed, Google's algorithms will also work to eliminate other similar explicit content that may arise in future searches. This comprehensive approach also includes the removal of duplicate images, ensuring that users have a more secure online presence and reducing the risk of such content resurfacing.

Emma Higham, a product manager at Google, emphasized the effectiveness of these protections, noting that similar measures have already been successful in managing other types of nonconsensual imagery. Higham stated, “These efforts are designed to give people added peace of mind, especially if they’re concerned about similar content about them popping up in the future.” This is particularly important in the age of rapidly evolving technology, where the threat of deepfake content can have severe implications for individuals' reputations and personal lives.

In addition to enhanced removal capabilities, Google is also modifying its search ranking algorithms to better manage queries that could lead to the display of explicit deepfake content. For instance, searches that explicitly seek deepfake images of real individuals will now be designed to surface “high-quality, non-explicit content” instead, such as relevant news articles or legitimate information. This is particularly relevant in light of incidents where explicit deepfakes of celebrities, like the recent AI-generated images of Taylor Swift, have circulated widely and caused significant concern.

Moreover, websites that frequently receive removal requests for fake explicit imagery will face demotion in Google Search rankings. This strategy is intended to discourage the proliferation of such harmful content and incentivize webmasters to remove or prevent the posting of explicit deepfake images on their platforms.

Google reports that earlier updates have already had a significant impact, reducing the visibility of explicit image results for queries specifically targeting deepfake content by over 70% this year alone. This remarkable figure highlights the effectiveness of Google's ongoing commitment to user safety and content management.

The tech giant is also working on innovations to differentiate between real explicit content and explicit fake content. This means that while deepfake imagery will be demoted in search results, legitimate images, such as consensual nude scenes featuring actors, can still be surfaced appropriately. This delicate balance is crucial in maintaining the integrity of content while protecting individuals from the potential harms of nonconsensual imagery.

These updates come as part of a larger strategy by Google to address the rising concerns regarding dangerous and explicit content online. Earlier this year, the company began banning advertisers from promoting deepfake porn services, reflecting a broader effort to curtail the exploitation of this technology for malicious purposes. Additionally, in 2022, Google expanded the types of "doxxing" information that can be removed from search results, and in August 2023, it implemented automatic blurring of sexually explicit imagery.

Together, these initiatives reflect Google's steadfast commitment to creating a safer and more secure online environment for its users. As deepfake technology continues to advance, the company recognizes the importance of adapting its policies and systems to protect individuals from emerging threats, ultimately fostering a digital landscape that prioritizes user safety and ethical standards.