Deepfakes have been a growing cause of concern in the media for several years, however the new influence of AI has made them front page news once again. Historical scandals include the deep-faking of pornographic material featuring celebrities (most famously Emma Watson) where the more recent scandals include Taylor Swift having deepfakes shared on X, formerly Twitter, where she is presented both sexually and in a “violently misogynist” way. Deep-faking is not only limited to pop culture figures, but also political figures – with audio replication of famous politicians’ voices becoming a popular meme concept, such as Barack Obama and Donald Trump playing Minecraft.
What is the impact of deepfakes?
Deepfakes have potential to have huge social and political impact, with concerns already being expressed around Kremlin involvement in their creation in order to influence foreign election results. Although social media platforms such as X have strict policy around not allowing deepfakes, many of the website and app algorithms are not advanced enough to remove the posts immediately, instead relying on manual reports of deep-faking.
How can authorities stop/police these incidents?
It is very challenging for authorities to manage the flow of AI production and to punish the perpetrators, due to the mass resharing and reposting of the images. Legislation in the UK is currently being reviewed following an open letter with signatures of 400 industry experts on AI listing concerns about pornography, child pornography, fraud and disinformation as key reasons to tighten laws surrounding the production and distribution of these deep-faked images. Although the individuals featured in the deepfakes are legally able to press charges against those who create and share the images, the platforms themselves go relatively unpunished for enabling distribution.
Legislation in the UK is currently being reviewed following an open letter with signatures of 400 industry experts on AI listing concerns about pornography, child pornography, fraud and disinformation as key reasons to tighten laws surrounding the production and distribution of these deep-faked images.
Are there good uses of deepfakes or are they always maliciously created?
Although there are many cases where deep-fakes are used to degrade, humiliate, and ruin careers, there are also cases where deep-fakes are used in a purely innocent manner such as in French video game studio company Drama’s case, where the players can scan their own faces and then be inserted into the game in order to see themselves within the game. The key AI challenges going forward will be finding the balance between restricting harmful material whist still enabling creatives to express themselves and use the technology to benefit wider society.
