Google addresses itself to non-consensual deepfakes
In connection with an abundant availability of non-consensual deepfakes in the internet space, a tech giant Google offers a way to cope with this issue. Of course, these manipulated images and videos, especially as they pertain to the negative cropping of female celebrities, have made it a rather intricate problem for search engines. It is also produced even when the viewer was not intentionally searching for this type of material.
Updating Filing Systems
Google’s reaction was made in the blog. It says that changes to the rating systems have been applied in order to reduce the number of representations of purely fictional pornography which are produced by hundreds of searches. Emma Higham, Google’s chief product officer stated that when one searches for specific terms associated with non-consensual deepfakes, the ranking system selects more serious content when available as opposed to fake images or videos.
An effective fight?
But these changes implemented by Higham have already helped in decreasing the vulnerability of people by a factor of 70% in deepfakes search results. However, in the same measure that Google is trying to demote legitimate deepfake content, it is not very easy sometimes to differentiate between the real and consensual material and that which is generated by AI with the help of a concerned actor.
Detecting and eliminating deepfakes
In order to overcome this issue, now Google is capable of identifying deepfake content by considering whether or not “a site’s page has been removed from Search based on the company’s policies.” Also, the firm is trying to make it easier for the victims who wish to unmask deepfakes on their own. After a deletion process, Google vowed to exclude all such related results as well as omit any of the duplicate entries.
Nevertheless, Google admits that there is a long way in this area and shares the company’s commitment to creating new solutions for deepfake victims who suffered from non-consensual intimate images. The intervention is happening two months after White House asked the Silicon Valley giants to limit distribution of graphic deepfake images.
