To tackle the rising issue of explicit “deepfake” images, Google has announced new measures to make it easier for individuals to remove AI-generated nudes of themselves from Search. This move aims to address the misuse of deepfake technology, which often involves superimposing someone’s face onto another person’s body, sometimes in sexually explicit contexts.

Deepfake technology can be both humorous and problematic. While it can create amusing content, such as placing Ron Swanson’s face over Wednesday Addams, it also poses significant risks. For instance, there have been instances of deepfakes falsely depicting Ukrainian President Volodymyr Zelenskyy instructing troops to surrender.

Initially used to create explicit images of celebrities, the accessibility of deepfake technology has led to its misuse against everyday individuals. Google’s new changes are designed to help remove such nonconsensual explicit images from Search results.

How Google’s New Measures Work

• Enhanced Removal Requests: When someone requests the removal of explicit nonconsensual deepfakes, Google’s systems will now aim to filter all related explicit searches.
• Duplicate Image Removal: If one explicit image is flagged, Google will also attempt to remove duplicates from Search results.
• Search Result Adjustments: Google will lower the ranking of explicit fake content and boost high-quality, nonexplicit content, reducing exposure to explicit results by 70%.

Steps to Remove AI-Generated Deepfakes

1. Request Removal: Follow the provided link to request the removal of explicit deepfakes.
2. Meet Requirements: Ensure the request meets the three criteria:
   • The person is identifiable in the imagery.
   • The imagery is fake and falsely depicts the person in a sexually explicit situation.
   • The imagery is distributed without the person’s consent.
3. Google Review: Google will review the request and respond with actions taken or explanations if the request doesn’t meet requirements.

The Broader Impact of AI Democratization

With the democratization of AI technology, especially following the release of ChatGPT in late 2022, issues like misinformation, cybercrime, and the spread of explicit fake images have surged. The ease of creating such content poses significant ethical and regulatory challenges.

Google’s Commitment to Quality

Google is committed to protecting users from harmful content by:

• Improving Ranking Systems: Lowering harmful content in Search results.
• Filtering Legitimate Content: Differentiating between legitimate and illegitimate explicit content by monitoring sites with frequent removal requests.

Final Thoughts

Google’s efforts to simplify the removal of explicit AI-generated deepfakes underscore the company’s dedication to user safety and content integrity. As AI technology evolves, so do the challenges, and Google’s proactive measures aim to mitigate these issues effectively.