By Yohanna D’Costa

Introduction

In February 2022, a surprising video of President Zelensky was shared on Twitter. It featured the president discussing a surrender to Russia. Of course, President Zelensky never said this. The slightly odd appearance of Zelensky led to the video being labelled a “deepfake.”  Deepfakes are becoming a regular feature of our digital lives. A 2023 study found that there has been a 550% rise in the creation of doctored images since 2019. With this increased exposure to deepfakes, New Zealand law should be examined. Are we equipped to reside in a falsified reality?

What are Deepfakes?

“Deepfakes” refer to false images, audio or videos created by artificial intelligence. The image in question has not been merely edited or retouched; it is manipulated in such a way that makes it exceedingly difficult to realise that it is not real.  Often, a deepfake will appear slightly unusual, which can lead to it being identified as an AI-produced image. But once a deepfake has been created, the damage is done.  A false version of an individual’s face or voice is then being used without their consent.

Deepfakes are not necessarily malicious and can be incredibly powerful. The Illinois Holocaust Museum and Education Center had an event where they used AI to create interactive images of 15 Holocaust survivors. Deepfake audio technology has been used to recreate John F. Kennedy’s voice and deliver the speech he was supposed to make in Dallas prior to his assassination.  Clearly, deepfake technology is having a transformative impact on the way we learn and experience the world. Yet, the global norm for deepfake usage is not so benign.

Why Should We Be Concerned?

The technology around deepfakes is becoming easier and more accessible. There are several programmes online that can help you create a deepfake. Essentially, anyone can make a deepfake with varying degrees of quality. Whilst there are rising political concerns, currently, deepfakes are predominantly used in nonconsensual pornography. One study found that 96% of deepfakes are pornographic videos of women. This was most notably seen with Taylor Swift. Earlier this year, deepfake explicit photos of the singer were circulating online. This prompted US politicians to call for new laws criminalising the creation of deepfakes. Deepfakes have become the newest form of digital sexual violence that women must navigate. Without regulation, deepfakes are an easy way for abusers to maintain distress and harm their victims. Some states have already recognised this. For example, England and Wales are going to make creating deepfake pornography a criminal offence.
 

Deepfakes have further sinister implications for our democracies and our political relations. In New Hampshire, voters received robocalls that used deepfake technology to mimic Joe Biden, discouraging Democrats from voting in their primary elections. Deepfakes could signal a new era for politics and manipulation. Just how destabilising this could be has yet to be determined. Easy access to deepfakes also has concerning possibilities for our justice system. Our evidence handling and criminal procedure are not prepared for people’s actions and words being deceptively synthesised.

New Zealand’s Current Law Around Deepfakes

No legislation explicitly addresses deepfake technology. The Harmful Digital Communications Act 2015 was made to address how technology can be used to harm others. It can cover issues such as online bullying and revenge porn. Revenge porn is a type of digital abuse where sexually explicit photos of an individual are shared without their consent. However, deepfakes do not meet the statutory definition of “intimate visual recording.” The main issue is that deepfakes do not technically feature the victim’s actual body.  This is little consolation to the victims of deepfakes who have not consented to their image being used in such a graphic manner. An amendment in March 2022 to The Harmful Digital Communications Act made sharing intimate visual recordings of another person (without their consent) an offence. However, a proposed amendment to include deepfake pornography in the act was not included. Despite the concerning implications of political deepfakes, they also remain relatively unregulated.  Sections 197 of the Electoral Act 1993 (interfering with or influencing voters) and 199A (publishing false statements to influence voters) could be relevant. Yet they only apply on polling day or in the two days before polling day.

The law itself may not be a comprehensive solution to the pervasiveness of deepfakes. Law reform would need to be accompanied by a range of tools. Social media regulation and public education can play a significant role in how the public interacts with deepfakes. Social media platforms should require deepfakes to be clearly labelled and identified. Deepfake pornography should be removed from the platform altogether.

Conclusion

Deepfake technology is rapidly progressing. New Zealand must try regulating it before it completely spirals out of our control. Creating and sharing deepfake pornography should be a crime. Social media should have authentication requirements for deepfake photos and audio. Of course, any attempt to regulate deepfakes wades into the precarious balance of freedom of expression.  Nonetheless, it is still a conversation worth having. How much do we value our honest images and words? How much do we value our reality?

The views expressed in the posts and comments of this blog do not necessarily reflect those of the Equal Justice Project. They should be understood as the personal opinions of the author. No information on this blog will be understood as official. The Equal Justice Project makes no representations as to the accuracy or completeness of any information on this site or found by following any link on this site. The Equal Justice Project will not be liable for any errors or omissions in this information nor for the availability of this information.