ABC News Verify spent only 100 Australian dollars to create a digital copy of the voice of federal senator Jackie Lambi, with her consent, using an easily accessible online platform.
This example clearly demonstrates the threat posed by artificial intelligence applications, creating synthetic copies of images and voices in the form of deepfakes, which are becoming cheaper and easier to use. These technologies not only endanger democratic processes, especially during elections, but also compromise the identity of individuals.
Australian laws currently inadequately protect individuals whose images or voices are digitally cloned without permission. Introducing the concept of “Personality Rights” may be a solution.
Modern deepfake technologies are capable of creating increasingly realistic content, making it difficult to distinguish between real and fake. Many individuals who were shown the clone of Senator Lambi’s voice by ABC initially did not realize it was fake.
This ease of creating unauthorized deepfakes and voice clones highlights the potential for misinformation and significant harm to individuals’ reputations.
In 2020, the impact of deepfakes was demonstrated when a political deepfake video featuring the then Prime Minister of Queensland, Anastasia Palatshuk, garnered about a million views on social media platforms.
Australian laws related to defamation, privacy, misuse of images, and consumer protection could theoretically apply to situations involving deepfake videos or audio clips. Individuals could also file complaints with the Office of the eSafety Commissioner.
While copyright laws could potentially protect a person’s image and voice, the nuances of their application are complex. Often, the individual whose likeness was cloned by an AI platform may not own the source material (such as images, videos, or audio recordings) and therefore may not have legal standing to pursue a lawsuit for infringement.