Deepfake Technology – Legal Challenges and Key Questions

1. INTRODUCTION

It was concerned that the title of the article which has been published in the Guardian in 2018. “You thought fake news was bad? Deep fakes are where truth goes to die”

Deepfakes are video clips and audio recordings that are fake but appear to be the real audio or visual content of a human person with the help of artificial intelligence. Deepfakes come in three general categories[1]:

Face Swap: Replacing one’s face with another’s in photo or video content.

Lip Sync: Making another appear to say something they are not in audio or video content.

Puppet Technique: Making one move in ways they do not naturally in video.

You’ve probably seen videos of Putin, Kim Jong-un or Lionel Messi saying or doing things that we don’t expect from them. It is deepfake. It may be joyful that videos about internationally recognized figures, about whom all kinds of things are written and drawn, and who are expected to have a high level of tolerance. However, these are the visible face of this technology. On the unseen side of the iceberg, there is content that violates personal rights and pornographic videos that specifically target women. The main problem with this issue is that deepfake technology is very convenient to use for scribble or defamation. This article aims to be an introduction in terms of the issues that this technology can create, especially in terms of personal rights and copyrights. It will be updated with the development of technology and the increase in legal issues.

2. PERSONAL RIGHTS

The fake contents have been producing by using the voice, image or motion of the persons through deepfake technology. The protection of personal rights is essential in law. According to the Turkish Civil Code (“TMK”), every attack to personal rights is unlawful. Naturally, there are exceptions for this rule in the legislations.

There is a discussion concerned if deepfake creators can benefit the freedom of speech as per as the First Amendment in USA. Especially, the protection of content within the scope of satire, parody and criticism of public figures within the scope of freedom of speech should be discussed. Turkish Constitutional Court declares that the freedom of speech covers all expressions like political, artistic, akademic, or commercial thougts and expressions. Therefore, according the Court satire or parody should be protected as a criticism type. The European Court of Human Rights has also emphasized that satire is an artistic expression and a form of social interpretation that presents the underlying reality in an exaggerated and distorted way, and that it naturally aims to provoke.

Accordingly, for Turkish law, the processing of the image of a public figure for criticism or artistic purposes; it should be noted that examples such as the use of this technology in order to make a visual work a parody can be considered lawful. However, in our opinion, this content has a very important difference from other satire, parody or criticism content. The general viewer may not be able to easily tell whether the content in question is real or not with the advancement of technology. Many people are likely to believe that these contents are real. For this reason, it may be the case that the courts put personal rights above freedom of speech. For instance, a column by a politician implying that he or she is being manipulated by another state can benefit from freedom of speech. However, an audio or video recording created with deepfake technology showing that this politician is receiving instructions from another state may not benefit from freedom of speech protection.

3. CRIMINAL LAW

The Turkish Penal Code (“TCK”) does not define a specific crime on this issue. In the doctrine, it is proposed to establish the “Crime of Disclosing, Disseminating, Making Accessible or Producing Sexually Explicit Images Without Consensuality” regarding sexual images produced with deepfake technology [1]. However, there is no such draft or regulation in Turkish law yet. Depending on the circumstances of the case, the courts may prosecute for sexual harassment, obscenity, insult, violation of privacy or unlawful use of personal data.

4. COPYRIGHT LAW

A person has a personal right over his appearance. There is also a right of personality to painting, which means the reflection and repetition of external appearance, and therefore paintings and portraits are primarily protected by the Civil Code as a right of personality. In Article 86 of the Law on Intellectual and Artistic Works (“FSEK”), paintings and portraits are protected by a special provision. Accordingly, persons whose images are used without permission can apply for legal remedies within the scope of Article 86 of the FSEK, as well as the TMK.

Another problem with deepfake content is the use of music or movies in these contents without permission from the authors. For example, a deepfake content in which Barack Obama sings a Taylor Swift song! There may also be those who claim that this content is a “work of work”.

In May 2020, WIPO published its Revised Issues Paper on Intellectual Property Policy and Artificial Intelligence. The main problems related to the subject in this policy were identified as follows:

• (i) Is copyright an appropriate vehicle for the regulation of deep fakes?

• (ii) Since deep fakes are created based on data that may be the subject of copyright, should the deep fake benefit from copyright?

• (iii) If deep fakes should benefit from copyright, to whom should the copyright in the deep fake belong?

• (iv) If deep fakes benefit from copyright, should there be a system of equitable remuneration for persons whose likenesses and “performances” are used in a deep fake?

WIPO states that the problems created by deepfake content are multidimensional. WIPO also states that deepfake content should not be rewarded with copyright protection if it portrays a person in a completely inconsistent or inaccurate way. In our opinion, an intellectual product that meets the conditions of being a work should have the title of a work, even if it violates the personal rights or intellectual property rights of the people it contains within the scope of FSEK. However, the owners of personal rights or intellectual property rights may prevent the author from exercising their rights and prevent the circulation of this content through the methods stipulated in the TMK and FSEK.

5. DATA PROTECTION LAW

In the Guideline on the Issues to be Considered in the Processing of Biometric Data published by the Turkish Personal Data Protection Board (“Board”), personal data arising from specific technical processing regarding the physical, physiological or behavioral characteristics of a natural person, such as facial images or dactyloscopic data, that enable or confirm the authentic identification of a natural person, are referred to as biometric data, citing the GDPR. Biometric data is also sensitive personal data. According to the Law on the Protection of Personal Data (“KVKK”), sensitive personal data, as a rule, cannot be processed without the explicit consent of the data subject.

We can say that the faces or voices of the people used in deepfake content are biometric data. Accordingly, persons whose personal data are processed unlawfully and used in deepfake content can also exercise their rights arising from the KVKK.

6. CONCLUSION

Many deepfake contents has been using in social media platforms nowadays. The legal problems concerned deepfake techonology will increase due to this technology will improve. For instance, there will be discussion point concerned the legal responsibilities of brands which produce deepfake contents on consumer law, or the control responsibility of the marketplaces which provide distribute this contents. Also, the election law issues will be another discussion point regarding this technology.

This problem, which is currently solved through general provisions, will begin to be addressed in a different way with possible Supreme Court decisions and legislative changes.