Deepfake technology is a frightening thing! While there are certainly good uses for this kind of technology, chaos thrives in unprecedented environments. Now, people everywhere have the ability to create fake images and videos of whomever they please.

Deepfake Definition

Deepfake technology applies neutral net simulation to massive datasets in order to create a synthetic output; hence the name ‘deepfake’.

How Do Deepfakes Work?

The power of artificial intelligence allows machines to learn what a source face looks like from multiple different angles. It will then transpose that face onto another source, whether it be a person or a robot, in a way that is similar to – and sometimes more realistic than- someone wearing a mask.

Deepfake Examples

One of the most well-known and controversial deepfake videos is of speaker Nancy Pelosi. In the video, Pelosi was altered to appear as though she was slurring her words. The video was posted on Facebook, and the company was criticized for refusing to take it down. In Facebook’s defense, the company claimed they had placed the video under its fact-checking process. Additionally, they reduced its reach on the social network as well, according to the New York Times.

 

Another political deepfake example happened recently. In what many are saying is an attempt to swing the demographic’s vote, images of Donald Trump with black voters were heavily circulated online. These photos even made it onto popular online news sources before people discovered that they were AI-generated images.

Deepfakes and Facebook

After being heavily scrutinized for refusing to remove the Nancy Pelosi video, Facebook is now banning videos that are heavily manipulated by artificial intelligence, i.e. deepfakes. On Monday, January 6th, 2020 Facebook posted a blog stating that the social network would “remove misleading manipulated media if it meets the following criteria:

• It has been edited or synthesized – beyond adjustments for clarity or quality – in ways that aren’t apparent to an average person and would likely mislead someone into thinking that a subject of the video said words that they did not actually say; and
• It is the product of artificial intelligence or machine learning that merges, replaces or superimposes content onto a video, making it appear to be authentic. This policy does not extend to content that is parody or satire, or video that has been edited solely to omit or change the order of words.”

In other words, Facebook is mostly concerned with the role deepfakes have in spreading fake news. Unsurprisingly, given that Facebook is classified as the No.1 platform for sharing false political stories, according to disinformation researchers.

Deepfakes and Snapchat

While Facebook is increasing their deepfake detection and working to ban malicious AI technology, Snapchat is working on their newest ‘deepfake’ features!

Snapchat is a picture and video sharing platform famous for its “snapchat lens” that can morph users’ faces into a younger version of themselves (“the baby filter”) or even appear as a dog on screen with a cartoon tongue that comes out of the user’s mouth when they open it.

They are currently developing a way for users to overlay their face onto a selection of pre-made scenes. TechCrunch reports that Snap Inc. quietly acquired an AI factory in Ukraine for a price believed to be around $166 million in order to create the new feature.

Deepfakes and Pornography

Unfortunately, deepfake pornography is nothing new. With the newest technology, though, it is easier than ever to sexually harass and defame people online. In fact, hundreds of Telegram app users have been victimized by a free deepfake bot! This bot replaces the clothed parts of women’s bodies with pornography, according to CNET. Despite the exponential increase in deepfakes, the United States government hadn’t commented on this pressing issue until Taylor Swift became the latest victim.

In recent artificial intelligence news, Taylor Swift AI-generated explicit photos went viral on the social media platform X. The Verge reports that White House press secretary Karine Jean-Pierre responded to the “alarming” news of Swift’s fake images and acknowledged how women are “the overwhelming targets of online harassment and also abuse”, but did not give any insight as to how the White House plans to address this issue.

One thing is clear though: the decision to downsize X’s content moderation team was not an advantageous one. The Taylor Swift deepfake AI images gained millions of views before the content was removed, and X has failed at restricting searches, according to CNN. Currently, the EU is investigating X’s content moderation practices, and this is sure to impact their findings. Hopefully, this incident will also, finally, prompt U.S. lawmakers to regulate non-consensual deepfake photography.

Are Deepfakes Ethical?

When deepfake technology was first introduced, it seemed relatively harmless because it was less realistic. The faces overlapping the videos were glitchy and most often used for comedy purposes. Now that technology is advancing, the line between entertainment and reality is quickly blurring. It’s very possible that this technology has the potential to spread fake news and defamation. In fact, one could argue that it has already has with the Nancy Pelosi video.

Deepfakes and the Law

What would be legal repercussions for creating deepfake videos that defame politicians, celebrities or businesses? Would it be considered defamation or would it be a different term?

Daniel Warner, a defamation attorney at RM Warner Law, weighs in:

“Depending on the particular facts and circumstances, along with how this technology is used under those facts and circumstances, the use of deepfakes may constitute defamation, false light and/or misappropriation of likeness. Also, it is possible that using deepfakes in a certain manner may rise to the level of criminal conduct. Criminal statutes are very broadly written. For example, under ARS 13-2008, ‘[a] person commits [the class 4 felony of] taking the identity of another person or entity if the person knowingly takes, purchases, manufactures, records, possesses or uses any personal identifying information or entity identifying information of another person or entity, including a real or fictitious person or entity, without the consent of that other person or entity, with the intent to obtain or use the other person’s or entity’s identity for any unlawful purpose or to cause loss to a person or entity whether or not the person or entity actually suffers any economic loss as a result of the offense . . . .’ Under this very broadly written statute, basic use of deepfakes may constitute this criminal offense. At the very least, it is important that people who use deepfakes make it abundantly clear that the use of a third party’s image is fictional. However, even with taking this precaution, the person may still be civilly, and even criminally, liable, depending on the facts and circumstances.”

Only time will tell if deepfakes become malicious enough to be considered criminal conduct, but it is likely that we will begin to see more and more defamation claims that stem from deepfakes.

Deepfake Help

If you are seeking the help of an experienced Internet defamation attorney, look no further than RM Warner Law. Our lawyers know exactly how to litigate cases for both public and private figures. Get in contact today.