Google Improves Search to Combat Deepfakes

Google has announced changes to its search algorithm and the process for removing unwanted content related to AI deepfakes. People who have been victims of explicit deepfakes can already ask Google to take down these fake images. Now, the company promises to make this process smoother.

Once Google finds a reported deepfake, it will block similar images from appearing in search results. This means people won’t have to repeatedly report the same or related fake photos.

The company says the problem of deepfakes is getting worse online. These fake photos are easy to make with new AI tools that anyone can use.

To combat this, Google will punish websites that post these harmful fake images. Sites that repeatedly share non-consensual deepfakes will be ranked lower in search results. Google says this method has worked for other harmful content and will help reduce fake photos online.

The company wants to stop fake explicit images from showing up in search results while still allowing real, legal adult content. However, Google admits it’s difficult to tell the difference between real and fake sexually explicit photos. This means the search engine might not always correctly identify these images.

Regardless of technical challenges, Google says that the changes made to Search have reduced deepfakes from resurfacing by up to 70%. Google said:

With these changes, people can read about the impact deepfakes are having on society, rather than see pages with actual non-consensual fake images.