How to remove deepfake yourself

In Russia and in the world, the creation and distribution of fake content is not yet legally regulated. Some social networks have developed their own procedures to block false information.

Deepfake is easy to create, but much more difficult to remove.

Let’s consider how this can be done on sites with the standard ability to remove fake content: Facebook, Instagram, Twitter and VKontakte.

Facebook

In early 2020, Facebook adopted its anti-fraudulent content policy. Moderators recognize him as such if:

  • it has been edited or synthesized in ways that are not obvious to the average person (other than quality adjustments). Such content is misleading into thinking that the subject said something in the video that he didn’t actually say;
  • video is a product of artificial intelligence or machine learning that combines, replaces, or superimposes content on a video to simulate its authenticity.

Most of the fakes are recognized by social network algorithms, but they are not perfect. To remove a fake, click on the sign with three dots in the upper right corner of the post. Then select the option “Get support or report the posting“.

Screenshot facebook.com

In the window that opens, select “False information“, then “Other“. After receiving the message, the site moderators will watch the content on their own.

You can follow the fate of the complaint in the Support Service. If the application is rejected, it is allowed to write again, explaining the situation in more detail. Link to Facebook’s Deepfake Policy.

Instagram

The social network is owned by Facebook, so the same rules apply here, but complaints are dealt with by their own support in a separate service.

To remove deepfake, just like on Facebook, click on the three-dot sign in the upper right corner of the post. Next, select the option “Complain” and “This is inappropriate content“. The further algorithm of actions is the same as for Facebook.

Screenshot instagram.com

The trouble is that if the complaint is accepted, Instagram does not remove the content, but only blocks its promotion using hashtags. But, if you talk to the support service, then she can meet halfway. Arguing the need to remove content, it is better to reinforce the rules for protecting the image of a citizen, supporting the appeal with the full text of the pre-trial claim.

Twitter

Back in 2019, the social network introduced rules for regulating posted content. According to them, content with the following features is recognized as fake:

  • a photo or video that has been significantly altered;
  • fraudulent content;
  • content may be a threat to public safety or may harm the user.

However, administrators do not remove the deepfake. First, it will be flagged as manipulative content. It will be possible to remove content if there are signs of deliberate spread of fake. True, Twitter does not indicate the criteria by which it determines this.

Screenshot help.twitter.com

Go to Twitter support, then you need to select “I want to report an intellectual property violation“. Let’s clarify: while deepfakes formally fall under the misuse of the image. You have not given permission to do so, so you have the right to request that the content be removed.

VKontakte

Open a page whose user, using your photos, impersonates you. Below the main profile photo (avatar), there is a button with three dots, click it. In the menu, select “Report page“.

After clicking on the button, a small window will open in which you need to select “Clone of my page (or my old page)“, and then click the Send button.

The application will be delivered to the administrators of the social network Vkontakte. The review is carried out in manual mode, so it will take some time. If the fact of using your photos is confirmed, the page will be frozen and deleted.
If there is no reaction, then there is another opportunity to remove it: write to the VK support service.

If the site does not have a special form for a deepfake complaint

If the site that published the deepfake does not have a policy on handling such content, then you need to write to Google. In the subsequent work with the appeal, it is better to rely on copyright. True, there are discrepancies in the law.

Formally, deepfake can be viewed as a result of intellectual activity. In this case, the copyright belongs by default to the person who created the fake.

The author of the result of intellectual activity is a citizen whose creative work has created such a result.

CC RF ARTICLE 1228

However, the content was created based on a photograph of a citizen who did not give permission to process it. Consequently, the derivative content appeared illegally. This is the hook that will remove deepfake. It is impossible to remove deepfake when the author of the fake was not obliged to ask permission from a citizen:

At the same time, there are cases when the author of the fake is not obliged to ask the permission of the citizen:

  • the photo was taken in public places, at events, frequently visited sites, concerts, etc .;
  • the photo shows a public figure or famous person;
  • the citizen posed for this photo for a fee.

If you follow the letter of the law, then the site administration needs to write a complaint:

Hello!
Your site contains content based on my photo. I have not given permission to anyone to process it or otherwise use it. The content is published at the link: site.ru/images/verydeepfake12.jpg
Please note that permission to use is required for photos, including those posted on social networks.
The publication of content falls within the scope of article 152.1. Civil Code of the Russian Federation “Protection of the image of a citizen” and Article 13.11. Administrative Code of the Russian Federation “Violation of the legislation of the Russian Federation in the field of personal data.”
I demand that you stop violating my rights by deleting the said content and its copies on your site. Otherwise, I will have to contact the law enforcement agencies.

Verified accounts

To distinguish real profiles from fake ones, there are special verified accounts on some social networks. Such accounts are marked with a special icon, which guarantees that the site administration has checked and the page really belongs to this person.

The rules for obtaining a confirmed status in most social networks are not clearly written, and are considered individually.

Fake social media accounts are not uncommon, and in March of this year we reported that deepfake celebrities were starting to appear on TikTok or Clubhouse.

One thought

Leave a Reply

Your email address will not be published. Required fields are marked *