Is that you? Deep dive into deepfakes part 2: Legal issues and regulatory landscape

In the first article of this three-part series we took a look at what deepfakes are and canvassed some of the ways they have been used. In this second instalment, we will examine the legal and regulatory framework which currently applies to deepfakes, both in Australia and overseas, and explore some of the legal issues which may arise.

Real (deep)fake legislation?

Australia has no specific legislation which addresses the potential misuse of deepfake technology, and we have yet to see a case concerning a deepfake reach the Australian judicial system. Other jurisdictions, however, have begun the process of legislating to address the potential for deepfakes to be misused.

Due to the growing number of deepfakes and their potential harmful uses, including with respect to politicians, the US states of California, Texas and Virginia have attempted to pass legislation regarding deepfakes. Former President Trump also passed national legislation in the US which required reporting on the foreign weaponization of deepfakes, including those targeted at disseminating disinformation relating to US elections and established a "Deepfakes Prize" competition to encourage research or commercialisation of deepfake detection technologies.

Both China and the European Union have also taken positive steps to combat harmful deepfakes, although China has also experimented with using deepfake news anchors to deliver news broadcasts.

A related issue that pervades the regulation of deepfakes is the myriad of practical difficulties with enforcing any legislative restrictions, particularly as:

• Detecting deepfakes is becoming harder as the technology becomes more advanced;
• Determining the creator of a deepfake can be difficult;
• Once the doubt has been raised, it is difficult to correct; and
• The speed with which these deepfakes can be disseminated far outstrips the speed with which their origin can be tracked.

While many have cried out for regulatory intervention, striking the right balance between competing rights and ensuring any legislation is agile enough to make an impact are challenges which need to be overcome in order to adequately address the concerns raised about this rapidly developing technology.

For better or worse, deepfakes currently occupy a place at a complex legal crossroads.

Intellectual property and deepfakes

One of the most obvious areas of law relevant to deepfakes is that of copyright, however, there are a number of complex issues which are likely to arise in this context. Firstly, it must be noted that only copyright owners are able to sue for infringement and the owner of the copyright is not always the person who is the subject of the copyright work. This means that while copyright may be one of the most obvious avenues for legal redress for those who have been the subject of a non-consensual deepfake, practically speaking, it may not be one that an aggrieved party is able to pursue personally. Assistance would be needed from the person or entity who owns the copyright, which in many cases where footage has been taken from movies or other large-scale productions would require major production companies agreeing to pursue copyright claims on behalf of those affected.

Another copyright issue that arises in the context of deepfakes relates to the ownership of the copyright in the deepfake itself. With the need for extensive input from AI and machine learning, attributing copyright ownership for media produced using a deepfake is beset by similar issues which arise in the context of AI generated artworks. At present, copyright protection is only afforded to works that have a human author.

Welcome to the party

As with many other rapidly developing technologies, the legal and regulatory frameworks which apply are lagging far behind deepfakes. With no specific legislation currently in force or contemplated by the Australian Parliament, businesses and individuals are left to navigate their way through the complex patchwork of existing laws, some of which offer more protection than others. As the recent US example has shown, however, implementing specific legislation to address the legal implications of deepfakes is not a simple or straightforward exercise, further complicated by the fact that the major players in this industry are based outside of Australia and so would likely escape the reach of any Australian legislation. Please contact Clayton Utz if you would like any assistance or advice relating to deepfake technology and its application.

Next edition: Recent examples, tips and tricks